Method and device for adjusting adaptive weight based on coefficient prediction, and equalization system

Abstract

The invention discloses a method and a device for adjusting an adaptive weight based on coefficient prediction, and an equalization system. The method comprises the following steps: acquiring a digital signal input to an adaptive equalizer, and predicting a tap coefficient target value of the adaptive equalizer according to the digital signal; acquiring a tap coefficient actual value of the adaptive equalizer for the digital signal; and configuring an adaptive weight for the adaptive equalizer according to the tap coefficient actual value and the tap coefficient target value, so that the tap coefficient actual value of the adaptive equalizer is iteratively updated based on the adaptive weight until the tap coefficient actual value converges. According to the method, the steady-state error can be reduced while the convergence speed is ensured.

Description

Technical Field

[0001] The present invention relates to the field of signal processing technology, and in particular to a method and device for adjusting adaptive weights based on coefficient prediction, and an equalization system. Background Art

[0002] In serial transmission systems (such as Ethernet, etc.), when data is transmitted in cables, the dispersion characteristics of the cables may cause signal distortion, which varies with the signal spectrum and the length of the medium loop. In order to ensure the integrity and accuracy of the signal, adaptive equalizers (such as forward feedback equalizers and decision feedback equalizers) are generally used to overcome signal distortion and inter-symbol interference during transmission and compensate for the received signal.

[0003] The adaptive equalizer can adjust its coefficients at any time based on the measurement of channel characteristics to adapt to changes in these characteristics, thereby effectively eliminating inter-symbol interference. An adaptive equalizer based on the LMS (Least Mean Square) algorithm is an effective solution. The core goal of the LMS algorithm is to minimize the mean square estimation error. It continuously adjusts the tap coefficients of the equalizer (the weight of the filter) in an iterative manner to optimize the equalizer's compensation effect on the received signal.

[0004] The adaptive weight is a key factor in the LMS algorithm, which determines the algorithm convergence speed and steady-state error. In the design of a constant-weight adaptive equalizer, it is impossible to optimize both the convergence speed and the steady-state error. A larger adaptive weight can speed up the convergence speed, but may lead to a larger steady-state error. A smaller adaptive weight can reduce the steady-state error, but the convergence speed will be slower. If the adaptive weight is too large, even when the algorithm is close to convergence, it may cause the weight update to be too large, causing the system to oscillate or even diverge. If the adaptive weight is too small, it may cause the algorithm to require more iterations to converge, thus wasting computing resources. Summary of the invention

[0005] The present invention provides a method and device for adjusting adaptive weights based on coefficient prediction, and an equalization system, which can reduce steady-state errors while ensuring convergence speed.

[0006] In one aspect of the present invention, a method for adjusting adaptive weights based on coefficient prediction is provided. The method comprises: obtaining a digital signal input to an adaptive equalizer to predict a target value of a tap coefficient of the adaptive equalizer according to the digital signal; obtaining an actual value of the tap coefficient of the adaptive equalizer for the digital signal; and configuring adaptive weights for the adaptive equalizer according to the actual value of the tap coefficient and the target value of the tap coefficient, so that the actual value of the tap coefficient of the adaptive equalizer is iteratively updated based on the adaptive weight until the actual value of the tap coefficient converges.

[0007] In another aspect of the present invention, a device for adjusting adaptive weights based on coefficient prediction is provided. The device includes: a coefficient prediction module configured to obtain a digital signal input to an adaptive equalizer to predict a target value of a tap coefficient of the adaptive equalizer according to the digital signal; and a weight configuration module configured to obtain an actual value of the tap coefficient of the adaptive equalizer for the digital signal, and configure adaptive weights for the adaptive equalizer according to the actual value of the tap coefficient and the target value of the tap coefficient, so that the actual value of the tap coefficient of the adaptive equalizer is iteratively updated based on the adaptive weight until the actual value of the tap coefficient converges.

[0008] In another aspect of the present invention, an equalization system is provided, which comprises the above-mentioned device for adjusting adaptive weights based on coefficient prediction and an adaptive equalizer electrically coupled to the device.

[0009] According to the technical solution of the present invention, a digital signal input to an adaptive equalizer is obtained to predict the target value of the tap coefficient obtained when the tap coefficient of the adaptive equalizer finally converges. At the same time, the actual value of the tap coefficient of the adaptive equalizer is obtained. At this time, the adaptive state of the current equalizer can be determined according to the actual value of the tap coefficient and the predicted target value. Therefore, the adaptive weight for adjusting the tap coefficient is dynamically configured according to the actual value and target value of the tap coefficient, so that the equalizer can dynamically adjust the tap coefficient according to the working state. Compared with the method of adjusting the tap coefficient by using a fixed adaptive weight, the dynamic adaptive weight of the present invention can select the optimal adjustment step size according to the state of the equalizer, thereby reducing the steady-state error while ensuring the convergence speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A flowchart of a method for adjusting adaptive weights based on coefficient prediction according to an embodiment of the present invention; Figure 2 is a schematic diagram of a shift processing method according to an embodiment of the present invention; Figure 3 A schematic diagram of changes in actual values ​​of tap coefficients and adaptive weights according to an embodiment of the present invention; Figure 4 A schematic diagram of changes in iterative updates of actual values ​​of tap coefficients according to an embodiment of the present invention; Figure 5 A block diagram of an adaptive equalization device based on coefficient prediction according to an embodiment of the present invention; Figure 6 FIG. 4 is an architecture diagram of a balancing system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0011] In order to explain the technical content, achieved objectives and effects of the present invention in detail, the following is an explanation in combination with the implementation modes and the accompanying drawings.

[0012] In the prior art, the adaptive weight is a key factor in the LMS algorithm, which determines the algorithm convergence speed and steady-state error. In the design of an adaptive equalizer with a constant weight, it is impossible to optimize the convergence speed and the steady-state error at the same time. A larger adaptive weight can speed up the convergence speed, but may lead to a larger steady-state error. A smaller adaptive weight can reduce the steady-state error, but the convergence speed will be slower. If the adaptive weight is too large, even when the algorithm is close to convergence, it may cause the weight update to be too large, causing the system to oscillate or even diverge. If the adaptive weight is too small, the algorithm may require more iterations to converge, thereby wasting computing resources.

[0013] In order to solve at least the above technical problems, the present disclosure provides a method for adjusting adaptive weights based on coefficient prediction. According to the present disclosure, a digital signal input to an adaptive equalizer is obtained to predict the target value of the tap coefficient obtained when the tap coefficient of the adaptive equalizer finally converges. At the same time, the actual value of the tap coefficient of the adaptive equalizer is obtained. At this time, the adaptive state of the current equalizer can be determined according to the actual value of the tap coefficient and the predicted target value, so the adaptive weight for adjusting the tap coefficient is dynamically configured according to the actual value and target value of the tap coefficient, so that the equalizer can dynamically adjust the tap coefficient according to the working state. In this way, compared with the method of adjusting the tap coefficient with a fixed adaptive weight, according to the embodiment of the present disclosure, the optimal adjustment step size can be dynamically selected according to the state of the equalizer, thereby reducing the steady-state error while ensuring the convergence speed.

[0014] In an adaptive equalization system equipped with a decision feedback equalizer, the adaptive adjustment of the tap coefficient is usually implemented based on the LMS algorithm. The formula of the LMS algorithm is specifically: W(n+1)=W(n)+μ×e(n)×X(n), where W(n+1) represents the tap coefficient after iterative update, W(n) represents the tap coefficient before iterative update, μ represents the adaptive weight, e(n) represents the decision error, and X(n) represents the input signal. In the adaptive process of the adaptive equalization system, the absolute value of the decision error e changes from large to small. When the absolute value of the decision error tends to 0, the tap coefficient W approaches a stable value, which is considered to be converged. The adaptive weight μ is a step factor used to adjust the speed of tap coefficient update. It can be seen that the adaptive equalizer can train the tap coefficient that can overcome inter-symbol interference according to the current cable transmission conditions. The tap coefficient should tend to be stable as the adaptive process proceeds, and in an ideal state, the tap coefficient should be equal to a certain estimated tap coefficient. The convergence state of the tap coefficient represents the convergence state of the adaptive equalization system. Therefore, according to the embodiment of the present disclosure, the tap coefficients predicted by the decision feedback equalizer are compared with the tap coefficients actually trained by the decision feedback equalizer to determine different adaptive stages of the adaptive equalizer, thereby adjusting the adaptive weight of the adaptive equalizer. When the actual value of the tap coefficient differs greatly from the target value of the tap coefficient, a larger adaptive weight is selected. As the adaptive process increases, the actual value of the tap coefficient becomes closer and closer to the target value of the tap coefficient, and the adaptive weight is gradually reduced.

[0015] Hereinafter, the technical solution according to the present disclosure will be described with reference to specific embodiments and in conjunction with the accompanying drawings.

[0016] Figure 1 1 is a flow chart showing a method 100 for adjusting adaptive weights based on coefficient prediction according to an embodiment of the present disclosure. Figure 1 The method 100 includes the following steps 102 to 106.

[0017] In step 102, a digital signal input to an adaptive equalizer is acquired to predict a target value of a tap coefficient of the adaptive equalizer according to the digital signal.

[0018] In some embodiments, predicting the tap coefficient target value of the adaptive equalizer according to the digital signal includes: determining attenuation data of the signal after line transmission according to the digital signal, calculating the line length of the transmission line according to the attenuation data, and predicting the tap coefficient target value according to the line length. In this way, since the attenuation state of the digital signal changes with the change of the line length, the inter-symbol interference that the adaptive equalizer needs to eliminate in the end can be predicted when the line length is determined, that is, the stable value reached when the tap coefficient of the adaptive equalizer finally converges.

[0019] In some embodiments, obtaining a digital signal input to the adaptive equalizer includes: obtaining a digital signal after the analog-to-digital converter performs conversion processing on the input signal. In this embodiment, predicting the tap coefficient target value of the adaptive equalizer based on the digital signal includes: predicting the attenuation state based on the digital signal, and predicting the first-stage tap coefficient of the decision feedback equalizer based on the attenuation state. In this way, the analog signal after cable attenuation is input to the analog-to-digital converter for sampling, and the distribution state of the inter-symbol interference of the signal is predicted based on the sampling result of the analog-to-digital converter, that is, the attenuation state of the signal is predicted. The tap coefficient target value of the decision feedback equalizer can be predicted based on the attenuation state.

[0020] In a specific scenario, the adaptive equalizer includes a decision feedback equalizer and a forward feedback equalizer, and the acquired digital signal and actual value of the tap coefficient are relevant data of the decision feedback equalizer.

[0021] In step 104, actual values ​​of tap coefficients of the adaptive equalizer for the digital signal are obtained.

[0022] In step 106, adaptive weights are configured for the adaptive equalizer according to the actual values ​​of the tap coefficients and the target values ​​of the tap coefficients, so that the actual values ​​of the tap coefficients of the adaptive equalizer are iteratively updated based on the adaptive weights until the actual values ​​of the tap coefficients converge.

[0023] In some embodiments, configuring the adaptive weight for the adaptive equalizer according to the actual value of the tap coefficient and the target value of the tap coefficient includes: generating a weight switching threshold according to the target value of the tap coefficient, and configuring the adaptive weight for the adaptive equalizer according to the actual value of the tap coefficient and the weight switching threshold. In this way, the tap coefficient target value is used as the coefficient value expected to be achieved by the adaptive training of the adaptive equalizer. Therefore, when the actual value of the tap coefficient is closer to the target value of the tap coefficient, it means that the actual value of the tap coefficient of the adaptive equalizer is closer to the convergence state. The weight switching threshold is determined based on the target value of the tap coefficient, so that the weight switching threshold divides the adaptive process of the adaptive equalizer into different stages. When the actual value of the tap coefficient reaches the weight switching threshold, it can be switched to the adaptive weight of the current stage, thereby realizing the dynamic configuration of the adaptive weight.

[0024] In some embodiments, the adaptive equalizer includes multiple equalizers, and when configuring the adaptive weights of the adaptive equalizers, the adaptive weights of all equalizers should remain consistent.

[0025] In some embodiments, generating a weight switching threshold according to the tap coefficient target value includes: loading the tap coefficient target value into a shift register so that the tap coefficient target value is shifted bit by bit in byte order, and the value after each shift processing is the value obtained after the tap coefficient target value is calculated based on the percentage of the current byte, and the value after each shift processing is used as the threshold adjustment value, and multiple weight switching thresholds are determined according to the tap coefficient target value and different threshold adjustment values. In this way, the tap coefficient target value is shifted through a shift register to obtain tap coefficient target values ​​of different percentages, thereby determining multiple weight switching thresholds according to a specific change rule. While ensuring the accuracy of the weight switching threshold division, the weight switching threshold can change with the change of the tap coefficient target value.

[0026] Figure 2 is a schematic diagram illustrating a shift processing method according to an embodiment of the present disclosure. Figure 2 In some embodiments, the tap coefficient target value is loaded into the 10 bits of the shift register, where 1 bit is the integer part and 9 bits are the decimal part. The tap coefficient target value is shifted to the right 3 times, each time by only 1 bit, and 2 -1 (0.5) tap coefficient target value, 2 -2 (0.25) tap coefficient target value and 2 -3 (0.125) tap coefficient target value, 2 -1 (0.5) tap coefficient target value, 2 -2 (0.25) tap coefficient target value and 2 -3 The tap coefficient target value (0.125) is the output threshold adjustment value. For example, if the tap coefficient target value is 0.8, the threshold adjustment values ​​output by the shift register are 0.5×0.8=0.4, 0.25×0.8=0.2, and 0.125×0.8=0.1.

[0027] Reference Figure 2 In some embodiments, the tap coefficient target value is taken as the minuend, the threshold adjustment value is taken as the subtrahend, and the difference between the tap coefficient target value and the threshold adjustment value is calculated, and the difference is the weight switching threshold. For example, if the tap coefficient target value is 0.8, the threshold adjustment values ​​are 0.4, 0.2, and 0.1, respectively, then the weight switching thresholds are 0.8-0.4=0.4, 0.8-0.2=0.6, and 0.8-0.1=0.7, respectively.

[0028] In some embodiments, configuring the adaptive weights for the adaptive equalizer according to the actual value of the tap coefficient and the weight switching threshold includes: determining multiple threshold intervals according to the multiple weight switching thresholds, and configuring the adaptive weights according to the comparison results of the multiple threshold intervals and the actual value of the tap coefficient. In this way, when the actual value of the tap coefficient is located in different threshold intervals, it means that the actual value of the tap coefficient is greater than / less than a certain weight switching threshold. Since the weight switching threshold is determined based on the target value of the tap coefficient, the actual value of the tap coefficient is located in different threshold intervals, which characterizes the difference between the actual value of the tap coefficient and the target value of the tap coefficient. According to the difference between the actual value of the tap coefficient and the target value, the adaptive state of the current adaptive equalizer can be determined, thereby selecting the optimal adaptive weight to balance the convergence speed and convergence accuracy.

[0029] In some embodiments, in order to avoid too frequent weight updates, two weight switching thresholds can be arbitrarily selected from multiple weight switching thresholds to determine the threshold interval. When 0 < actual value of tap coefficient ≤ first weight switching threshold, it is considered that the equalizer is in the initial stage, and the adaptive weight of the initial stage is configured. When the first weight switching threshold < actual value of tap coefficient ≤ second weight switching threshold, it is considered that the equalizer is in the transition stage, and the adaptive weight of the transition stage is configured. When the second weight threshold < actual value of tap coefficient, it is considered that the equalizer is in the stable stage, and the adaptive weight of the stable stage is configured. In a specific application scenario, the tap coefficient target value predicted by the decision feedback equalizer is 0.8, and the weight switching thresholds generated based on the tap coefficient target value are 0.4, 0.6 and 0.7, respectively. 0.4 and 0.7 are selected from the weight switching thresholds to determine the threshold interval. Three threshold intervals can be determined based on the weight switching thresholds of 0.4 and 0.7, the first threshold interval is less than or equal to 0.4, the second threshold interval is greater than 0.4 and less than or equal to 0.7, and the third threshold interval is greater than 0.7.

[0030] In some embodiments, configuring the adaptive weight according to the comparison results of the multiple threshold intervals and the actual value of the tap coefficient includes: comparing the actual value of the tap coefficient with the multiple threshold intervals respectively to determine the threshold interval to which the actual value of the tap coefficient belongs, and configuring the adaptive weight as a target weight corresponding to the threshold interval. In this way, corresponding target weights can be pre-configured under different threshold intervals, and when the actual value of the tap coefficient is in the same threshold interval, the same target weight is used to iteratively update the tap coefficient, thereby avoiding training errors caused by frequent updates of the adaptive weight.

[0031] In a specific application scenario, the target weight corresponding to the first threshold interval is 2 -4 , the target weight corresponding to the second threshold interval is 2-6 , the target weight corresponding to the third threshold interval is 2 -8 If the actual value of the tap coefficient is within the first threshold interval, that is, the actual value of the tap coefficient is less than or equal to 0.4, the adaptive weight of the adaptive equalizer is configured to 2. -4 If the actual value of the tap coefficient is within the second threshold interval, that is, the actual value of the tap coefficient is greater than 0.4 and less than or equal to 0.7, the adaptive weight of the adaptive equalizer is configured to 2. -6 If the actual value of the tap coefficient is within the third threshold interval, that is, the actual value of the tap coefficient is greater than 0.7, the adaptive weight of the adaptive equalizer is configured to 2. -8 It should be noted that the predicted value of the tap coefficient and the actual value of the tap coefficient in the present invention are both absolute values ​​of the tap coefficient.

[0032] In some embodiments, until the actual value of the tap coefficient converges includes: if it is detected that the coefficient difference between the actual value of the tap coefficient and the target value of the tap coefficient is less than a preset value, then stop adjusting the adaptive weight. In this way, ideally, the actual value of the tap coefficient should be equal to the target value of the tap coefficient, but because the target value of the tap coefficient is a predicted result, there may be a certain error between the predicted result and the result of the actual convergence of the tap coefficient, so when the coefficient difference between the two is less than the preset value, it means that the tap coefficient has reached a converged state.

[0033] Figure 3 is a schematic diagram showing changes in actual values ​​of tap coefficients and adaptive weights according to an embodiment of the present disclosure. Figure 3 The horizontal coordinate represents the adaptive period, and the vertical coordinate represents the actual value of the tap coefficient. Figure 3 In a specific application scenario, the tap coefficient of the decision feedback equalizer is a negative number, the target value of the tap coefficient predicted by the decision feedback equalizer is 0.8, the initial value of the tap coefficient of the decision feedback equalizer is 0, and different adaptive weights are configured in different threshold intervals to adjust the actual value of the tap coefficient so that the actual value of the tap coefficient gradually approaches the predicted value of the tap coefficient 0.8.

[0034] Figure 4 is a schematic diagram showing changes in iterative updating of actual values ​​of tap coefficients according to an embodiment of the present disclosure. Figure 4 The horizontal coordinate represents the signal period, the vertical coordinate represents the signal amplitude, and the black dots represent the tap coefficients. When the black dots are located on the signal waveform, it means that the tap coefficients converge, so that the adaptive equalizer converges. Figure 4 (a) in the figure indicates that the actual value of the tap coefficient is 0, the difference between the actual value of the tap coefficient and the target value of the tap coefficient is 0.8, and the equalizer is in the initial period of the adaptive stage. Figure 4(b) in the figure indicates that the actual value of the tap coefficient is 0.4, and the difference between the actual value of the tap coefficient and the target value of the tap coefficient is 0.4, and the equalizer is in a transition period of the adaptive stage. Figure 4 (c) in the figure indicates that the actual value of the tap coefficient is 0.78, and the difference between the actual value of the tap coefficient and the target value of the tap coefficient is 0.02. The equalizer is in a stable period of the adaptive stage. Figure 4 (d) in the figure indicates that the actual value of the tap coefficient is 0.8, the difference between the actual value of the tap coefficient and the target value of the tap coefficient is 0, and the equalizer is in a converged state.

[0035] According to another aspect of the present invention, Figure 5 2 is a block diagram showing an adaptive equalization device 200 based on coefficient prediction according to an embodiment of the present invention. Figure 5 The device 200 includes a coefficient prediction module 202 and a weight configuration module 204.

[0036] The coefficient prediction module 202 is configured to obtain a digital signal input to the adaptive equalizer to predict a tap coefficient target value of the adaptive equalizer according to the digital signal.

[0037] The weight configuration module 204 is configured to obtain the actual value of the tap coefficient of the adaptive equalizer for the digital signal, and to configure the adaptive weight for the adaptive equalizer according to the actual value of the tap coefficient and the target value of the tap coefficient, so that the actual value of the tap coefficient of the adaptive equalizer is iteratively updated based on the adaptive weight until the actual value of the tap coefficient converges.

[0038] According to yet another aspect of the present invention, Figure 6 3 is a diagram showing an architecture of a balancing system 300 according to an embodiment of the present invention. Figure 6 The system 300 includes an adaptive equalizer 302 and the above-mentioned coefficient prediction-based adaptive equalization device 200, and the adaptive equalizer 302 is electrically coupled to the device 200.

[0039] In some embodiments, the system 300 may further include an analog-to-digital converter 304. The output of the analog-to-digital converter 304 is electrically coupled to the adaptive equalizer 302 and the coefficient prediction-based adaptive equalization apparatus 200, respectively.

[0040] In summary, the present invention provides a method and device for adjusting adaptive weights based on coefficient prediction, and an equalization system, which predicts the tap coefficient target value to which the tap coefficient of the adaptive equalizer finally converges according to the input signal, and generates different weight switching thresholds according to the tap coefficient target value. When the actual value of the tap coefficient of the adaptive equalizer is in the threshold interval determined by the weight switching threshold, the adaptive weight of the adaptive equalizer is configured to the target weight corresponding to the interval. Compared with the method of fixing adaptive weights, the present invention dynamically configures weights based on the real-time changes of the tap coefficients, and can select a larger weight to speed up the convergence speed when the tap coefficient is far away from the final convergence value, and select a smaller weight to control the adjustment accuracy when the tap coefficient is close to the final convergence value, thereby reducing the steady-state error while ensuring the convergence speed.

[0041] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's specification and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for adjusting adaptive weights based on coefficient prediction, characterized in that: include: Acquire a digital signal input to an adaptive equalizer to predict a tap coefficient target value of the adaptive equalizer according to the digital signal; Obtaining actual values ​​of tap coefficients of the adaptive equalizer for the digital signal; as well as Adaptive weights are configured for the adaptive equalizer according to the actual values ​​of the tap coefficients and the target values ​​of the tap coefficients, so that the actual values ​​of the tap coefficients of the adaptive equalizer are iteratively updated based on the adaptive weights until the actual values ​​of the tap coefficients converge.

2. The method according to claim 1, characterized in that Predicting the tap coefficient target value of the adaptive equalizer according to the digital signal comprises: Determine attenuation data of the signal after it is transmitted through the line according to the digital signal; Calculating the line length of the transmission line according to the attenuation data; and The tap coefficient target value is predicted according to the line length.

3. The method according to claim 1, characterized in that Configuring adaptive weights for the adaptive equalizer according to the tap coefficient actual value and the tap coefficient target value includes: generating a weight switching threshold value according to the tap coefficient target value; and The adaptive weights for the adaptive equalizer are configured according to the tap coefficient actual values ​​and the weight switching threshold.

4. The method according to claim 3, characterized in that Generating a weight switching threshold according to the tap coefficient target value includes: Loading the tap coefficient target value into a shift register, so that the tap coefficient target value is shifted bit by bit in byte order, and the value after each shift process is the value obtained after the tap coefficient target value is calculated as a percentage based on the current byte; and The value after each shifting process is used as a threshold adjustment value, and a plurality of weight switching thresholds are determined according to the tap coefficient target value and different threshold adjustment values.

5. The method according to claim 4, characterized in that Configuring the adaptive weights for the adaptive equalizer according to the actual values ​​of the tap coefficients and the weight switching threshold comprises: Determining a plurality of threshold intervals according to the plurality of weight switching thresholds; and The adaptive weights are configured according to comparison results between the multiple threshold intervals and the actual values ​​of the tap coefficients.

6. The method according to claim 5, characterized in that Configuring the adaptive weight according to the comparison results of the multiple threshold intervals and the actual values ​​of the tap coefficients comprises: Comparing the actual value of the tap coefficient with the multiple threshold intervals respectively to determine the threshold interval to which the actual value of the tap coefficient belongs; and The adaptive weight is configured as a target weight corresponding to the threshold interval.

7. The method according to claim 1, characterized in that Until the actual value of the tap coefficient converges, it includes: If it is detected that the coefficient difference between the actual value of the tap coefficient and the target value of the tap coefficient is smaller than a preset value, the adjustment of the adaptive weight is stopped.

8. The method according to claim 1, characterized in that: Acquiring a digital signal input to the adaptive equalizer includes: acquiring a digital signal converted by an analog-to-digital converter on the input signal, Predicting the tap coefficient target value of the adaptive equalizer according to the digital signal includes: predicting an attenuation state according to the digital signal, and predicting a first-stage tap coefficient of a decision feedback equalizer according to the attenuation state.

9. A device for adjusting adaptive weights based on coefficient prediction, characterized in that: include: A coefficient prediction module, configured to obtain a digital signal input to the adaptive equalizer to predict a tap coefficient target value of the adaptive equalizer according to the digital signal; as well as The weight configuration module is configured to obtain the actual value of the tap coefficient of the adaptive equalizer for the digital signal, and configure the adaptive weight for the adaptive equalizer according to the actual value of the tap coefficient and the target value of the tap coefficient, so that the actual value of the tap coefficient of the adaptive equalizer is iteratively updated based on the adaptive weight until the actual value of the tap coefficient converges.

10. A balancing system, characterized in that: include: The device according to claim 9; as well as An adaptive equalizer is electrically coupled to the device.

Images