Signal equalization device, signal equalization method and serial communication receiving system

Abstract

The invention relates to the technical field of electronic communication, and discloses a signal equalization device, a signal equalization method and a serial communication receiving system, and the device comprises a fixed tap feed-forward equalizer, a floating tap feed-forward equalizer and an adaptive engine. The fixed tap provides stable balance for a main peak area of channel impulse response, the floating tap dynamically adapts to the change of a secondary peak area, real-time feedback adjustment of the adaptive engine is combined, and the signal integrity is remarkably improved through the collaborative architecture of the fixed tap and the floating tap, so that the signal quality is improved. The problem that the signal integrity is reduced due to channel reflection and intersymbol interference in high-speed serial communication is solved.

Description

Technical Field

[0001] This application relates to the field of electronic communication technology, specifically to signal equalization devices, signal equalization methods, and serial communication receiving systems. Background Technology

[0002] In high-speed serial communication systems, as data transmission rates develop towards 56Gbps, 112Gbps, and even higher, signal integrity faces severe challenges. Factors such as insertion loss, reflection effects, and multipath interference in the channel lead to increasingly prominent inter-symbol interference problems, severely restricting system transmission performance. Especially in applications such as optical modules and high-speed backplanes, feedforward equalization technology, as a key means of compensating for channel loss, directly affects the reliability of the communication system.

[0003] Fixed-tap equalization schemes in related technologies have significant shortcomings when processing high-speed signals. When dealing with complex and variable channel environments, these schemes, due to their fixed tap positions, struggle to adapt to dynamic changes in channel reflection points, resulting in limited suppression of inter-symbol interference. Furthermore, their equalization capability is limited by preset tap coefficients and cannot adaptively adjust according to real-time channel conditions, making it difficult to maintain stable signal quality in high-speed data transmission scenarios.

[0004] While adjustable tap schemes in related technologies offer improvements in flexibility, they still suffer from numerous technical bottlenecks. These schemes are limited in tap adjustment accuracy by factors such as clock rate, making fine-grained compensation difficult. More importantly, these schemes perform poorly in handling signal integrity degradation caused by channel reflections and inter-symbol interference in high-speed serial communication. Particularly when dealing with long-tailed interference, they struggle to effectively compensate for interference components far from the main cursor, leading to unstable signal eye diagram quality, impacting bit error rate performance, and severely hindering further improvements in system performance. Summary of the Invention

[0005] This application provides a signal equalization device, a signal equalization method, and a serial communication receiving system to solve the problem of signal integrity degradation caused by channel reflection and inter-symbol interference in high-speed serial communication.

[0006] In a first aspect, this application provides a signal equalization device, comprising: a fixed-tap feedforward equalizer for equalizing an input digital signal to compensate for inter-symbol interference in the main peak region of the channel impulse response; a floating-tap feedforward equalizer for receiving the input digital signal, the floating-tap feedforward equalizer including a multiplexer and multiple tap groups, each tap group corresponding to a secondary peak region in the channel impulse response, the multiplexer being configured to dynamically switch activated tap groups based on the time-domain characteristics of the channel impulse response to compensate for inter-symbol interference in the secondary peak region; and an adaptive engine connected to the fixed-tap feedforward equalizer and the floating-tap feedforward equalizer, respectively, and receiving feedback signals from a decision unit, configured to generate error information characterizing signal quality based on a reference and the feedback signals, and to adjust the tap coefficients of the fixed-tap feedforward equalizer and the tap coefficients of the activated tap groups in the floating-tap feedforward equalizer according to the error information.

[0007] Beneficial effects: By employing a collaborative architecture of fixed and floating taps, the lack of flexibility in fixed-tap FFEs in related technologies is resolved. Fixed taps provide stable equalization for the main peak region of the channel impulse response, while floating taps dynamically adapt to changes in the secondary peak region. Combined with real-time feedback adjustments from the adaptive engine, signal integrity is significantly improved. This design avoids the shortcomings of traditional FFEs in terms of insufficient equalization performance in complex channel scenarios, achieves on-demand resource allocation, and reduces hardware redundancy.

[0008] In one alternative implementation, each tap group in the floating tap feedforward equalizer contains at least two taps, and the multiplexer is configured to uniformly activate or disable all taps within the same tap group.

[0009] Beneficial effects: Unified control through grouping simplifies the management logic of floating taps and reduces the complexity of control circuits. Each tap group dynamically switches as a whole unit, avoiding the timing challenges and hardware overhead caused by controlling each tap individually, and improving system response speed.

[0010] In one alternative implementation, the secondary peak region is the peak region in the channel impulse response whose amplitude is lower than the main peak but higher than the noise threshold.

[0011] Beneficial effects: Clearly defining the range of the secondary peak region ensures the accuracy of floating tap compensation. Threshold limiting avoids miscompensation caused by noise interference, improving equalization efficiency.

[0012] In one alternative implementation, the adaptive engine is configured to process the error information using a least mean square algorithm to adjust the tap coefficients of the fixed-tap feedforward equalizer and the floating-tap feedforward equalizer.

[0013] Beneficial effects: The introduction of the LMS algorithm enables rapid convergence and stable optimization of the tap coefficients, overcoming the slow adaptive behavior of related technologies. By iteratively minimizing the error, the algorithm ensures that the equalization parameters track channel changes in real time, improving the system's reliability in dynamic environments.

[0014] In an optional implementation, the adaptive engine is further configured to: perform initial channel impulse response analysis when the error level represented by the feedback signal exceeds a preset threshold, identify the largest group region of non-peak regions in the channel impulse response through real-time online scanning, and determine the initial tap group position of the floating tap feedforward equalizer based on the identification result.

[0015] Beneficial effects: The threshold-triggered initialization mechanism ensures the system's rapid adaptation during channel mutations, which helps solve the problem of cold start adaptation delay, enables floating taps to quickly locate key reflection points, and improves system robustness.

[0016] In one alternative implementation, the apparatus further includes a microcontroller connected to the adaptive engine and configured to control the adaptive engine to perform channel impulse response analysis during the initialization phase to determine the initial tap group position of the floating tap feedforward equalizer.

[0017] Beneficial effects: The introduction of the MCU enables fine control of the initialization process, and the accuracy of tap position optimization is ensured through initial analysis managed by firmware.

[0018] Secondly, this application provides a serial signal receiving system, comprising: an analog front-end for receiving and preprocessing analog signals; an analog-to-digital converter connected to the analog front-end for converting the preprocessed analog signals into digital signals; a clock data recovery unit connected to the analog-to-digital converter for extracting a clock signal from the digital signals and outputting synchronized data signals; a signal equalization device of the first aspect or any corresponding embodiment thereof connected to the clock data recovery unit for equalizing the synchronized data signals; a decision unit connected to the signal equalization device to receive the equalized data signals and connected to the clock data recovery unit to receive the clock signal, for making a decision on the equalized data signals using the clock signal and outputting digital data; and a decision feedback equalizer connected to the decision unit for performing feedback equalization based on the decision result to eliminate inter-code interference.

[0019] Thirdly, a signal equalization method is provided, the signal equalization method being implemented based on the signal equalization device of the first aspect or any corresponding embodiment thereof, comprising: equalizing the input digital signal through the fixed-tap feedforward equalizer to compensate for inter-symbol interference in the main peak region of the channel impulse response; equalizing the input digital signal through the floating-tap feedforward equalizer to compensate for inter-symbol interference in the secondary peak region of the channel impulse response; inputting the digital signal equalized by the fixed-tap feedforward equalizer and the floating-tap feedforward equalizer to a decision unit to obtain a feedback signal; generating error information characterizing signal quality through the adaptive engine based on a reference benchmark and the feedback signal; and adjusting the tap coefficients of the fixed-tap feedforward equalizer and the tap coefficients of the activated tap group in the floating-tap feedforward equalizer according to the error information through the adaptive engine.

[0020] In one optional implementation, the step of equalizing the input digital signal through the floating tap feedforward equalizer to compensate for inter-symbol interference in the secondary peak region of the channel impulse response includes: The active tap group is dynamically switched by a multiplexer based on the time-domain characteristics of the channel impulse response.

[0021] In one optional implementation, the step of adjusting the tap coefficients based on the error information includes: processing the error information using a least mean square algorithm to adjust the tap coefficients of the fixed tap feedforward equalizer and the floating tap feedforward equalizer.

[0022] In an optional implementation, the signal equalization method further includes: when the error level represented by the feedback signal exceeds a preset threshold, performing an initial channel impulse response analysis, identifying the largest group region of non-peak regions in the channel impulse response through real-time online scanning, and determining the initial tap group position of the floating tap feedforward equalizer based on the identification result. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0024] Figure 1 This is a schematic diagram of a signal equalization device according to an embodiment of this application; Figure 2 This is a schematic diagram of a serial signal receiving system according to an embodiment of this application; Figure 3This is a schematic diagram of the channel ISI corresponding to fixed taps and floating taps according to embodiments of this application; Figure 4 This is a diagram showing the test results of the floating TAP position and corresponding coefficients according to the FT-FFE algorithm of the embodiments of this application; Figure 5 This is a schematic flowchart of a signal equalization method according to an embodiment of this application; Figure 6 This is a schematic diagram illustrating the specific calculation process of ISI and FT-FFE according to embodiments of this application. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] It is understood that before using the technical solutions disclosed in the various embodiments of this application, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this application in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0027] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0028] As data transmission rates increase to 56Gbps, 112Gbps, and even higher, the application scenarios of signal links become more complex, leading to different reflection tap positions. Fixed-tap feed-forward equalization (FFE) structures lack flexibility and cannot optimize the equalization effect when facing certain specific channel characteristics, resulting in insufficient margin after signal equalization. Adjustable-tap schemes in related technologies perform poorly in addressing signal integrity degradation caused by channel reflections and inter-symbol interference in high-speed serial communication.

[0029] Reference Figure 1As shown in the embodiment of this application, a signal equalization device is provided, including a fixed-tap feedforward equalizer, a floating-tap feedforward equalizer, and an adaptive engine. The fixed-tap feedforward equalizer and the floating-tap feedforward equalizer both receive input digital signals, while the adaptive engine is connected to the fixed-tap feedforward equalizer and the floating-tap feedforward equalizer respectively and receives feedback signals from the decision unit.

[0030] Specifically, a fixed-tap feedforward equalizer is used to equalize the input digital signal to compensate for inter-symbol interference in the main peak region of the channel impulse response.

[0031] The floating tap feedforward equalizer receives the input digital signal. The floating tap feedforward equalizer includes a multiplexer and multiple tap groups. Each tap group corresponds to a sub-peak region in the channel impulse response. The multiplexer is configured to dynamically switch the activated tap group based on the time-domain characteristics of the channel impulse response to compensate for inter-symbol interference in the sub-peak region.

[0032] An adaptive engine is connected to both a fixed-tap feedforward equalizer and a floating-tap feedforward equalizer, and receives feedback signals from the decision unit. It is configured to generate error information characterizing signal quality based on a reference benchmark and feedback signals, and adjust the tap coefficients of the fixed-tap feedforward equalizer and the tap coefficients of the activated tap groups in the floating-tap feedforward equalizer according to the error information.

[0033] It should be noted that, in this embodiment, inter-symbol interference (ISI) refers to the mutual interference phenomenon between adjacent symbols caused by channel characteristics. The multiplexer can select one output from multiple inputs based on control signals, and in this scheme, it is used to achieve dynamic switching of tap groups.

[0034] In some specific examples, the reference point can be an ideal constellation point or a training sequence, and the calculated error signal includes mean square error (MSE), bit error rate (BER), signal-to-noise ratio (SNR), and eye diagram opening. Among these, the bit error rate is an important indicator of the reliability of a digital communication system, representing the proportion of erroneous bits to the total number of transmitted bits. In practical implementation, the tap positions of the fixed-tap feedforward equalizer can be preset within the range of [-2UI, +2UI], covering key interference points in the main peak region. Here, the unit interval (UI) is the basic unit for measuring signal timing, with one UI corresponding to the duration of one symbol period. The number of tap groups in the floating-tap feedforward equalizer can be dynamically configured according to channel complexity; for example, two groups can be used in simple channels, while four groups can be expanded in complex channels to balance resource overhead and performance. The error information generation cycle of the adaptive engine can be set to once every 10ms, ensuring real-time performance while avoiding excessive computational load. This design achieves a balance between versatility and adaptability through hierarchical management of hardware resources.

[0035] In these embodiments, a collaborative architecture of fixed and floating taps addresses the lack of flexibility in fixed-tap FFEs in related technologies. Fixed taps provide stable equalization for the main peak region of the channel impulse response, while floating taps dynamically adapt to changes in the secondary peak region. Combined with real-time feedback adjustments from an adaptive engine, signal integrity is significantly improved. This design avoids the shortcomings of traditional FFEs in terms of insufficient equalization performance in complex channel scenarios, achieves on-demand resource allocation, and reduces hardware redundancy.

[0036] Optionally, in some embodiments of this application, each tap group in the floating tap feedforward equalizer contains at least two taps, and the multiplexer is configured to uniformly activate or disable all taps within the same tap group.

[0037] It should be noted that a tap refers to the basic processing unit in an equalizer. Each tap contains a delay unit and a multiplier, used for weighting the signal. Activation or disabling control refers to the operation of controlling the tap's operating state through an enable signal. In the disabled state, the tap does not participate in signal processing to reduce power consumption.

[0038] In more specific examples, each tap group can contain three taps, corresponding to the front, middle, and rear positions of the secondary peak region, respectively. The multiplexer controls the activation state of the entire tap group through a set of enable signals. This unified control mechanism reduces the number of control lines and lowers wiring complexity; at the same time, the tap spacing within the tap group can be set to 1UI to ensure continuous coverage of secondary peak interference. In specific circuit implementations, the multiplexer can use a tree structure to optimize timing and improve switching speed.

[0039] In these embodiments, unified control through grouping simplifies the management logic of floating taps and reduces the complexity of the control circuitry. Each tap group dynamically switches as a whole unit, avoiding the timing challenges and hardware overhead of controlling each tap individually, and improving system response speed.

[0040] Optionally, in some embodiments of this application, the secondary peak region is a peak region in the channel impulse response whose amplitude is lower than the main peak but higher than the noise threshold.

[0041] It should be noted that the channel impulse response is a characteristic function describing the channel's response to a unit impulse signal, reflecting the channel's time-domain characteristics. The noise threshold is a threshold value set according to the system noise level, used to distinguish between valid signals and noise interference.

[0042] In some more specific examples, the noise threshold can be set below -20dB to effectively distinguish between real interference and random noise; the identification of secondary peak regions can be accomplished through the initial scan of the adaptive engine, for example, by using a sliding window algorithm to detect peak points whose amplitude exceeds the threshold but is 10dB below the main peak. In actual channels, secondary peak regions may be caused by connector reflections or multipath effects, and their positions may drift with changes in temperature or frequency, thus requiring dynamic tracking.

[0043] In these embodiments, the range of the secondary peak region is clearly defined, ensuring the accuracy of floating tap compensation. Threshold limitation avoids erroneous compensation caused by noise interference, improving equalization efficiency.

[0044] Optionally, in some embodiments of this application, the adaptive engine is configured to process error information using a least mean square algorithm to adjust the tap coefficients of the fixed tap feedforward equalizer and the floating tap feedforward equalizer.

[0045] It should be noted that the Least Mean Square (LMS) algorithm is a commonly used adaptive filtering algorithm that optimizes the filter coefficients by iteratively minimizing the mean square error. The tap coefficients refer to the weighted values ​​of each tap in the equalizer, which determine the degree of correction that tap makes to the signal.

[0046] In some more specific examples, the step size parameter of the LMS algorithm can be configured in the range of 0.001-0.01. A larger step size (e.g., 0.01) is used initially for rapid convergence, and then a smaller step size (e.g., 0.001) is switched to improve accuracy after stabilization. The algorithm's iteration cycle is synchronized with the data symbol rate to ensure timely coefficient updates. Furthermore, a momentum term can be introduced to avoid local optima and enhance algorithm stability.

[0047] In these embodiments, the introduction of the LMS algorithm enables rapid convergence and stable optimization of the tap coefficients, addressing the slow adaptive behavior of related technologies. This algorithm iteratively minimizes the error, ensuring that the equalization parameters track channel changes in real time, thus improving the system's reliability in dynamic environments.

[0048] Optionally, in some embodiments of this application, when verifying the effectiveness of this application, the adaptive engine is also configured to calculate the bit error rate by counting the number of erroneous bits per unit time, and / or calculate the eye opening by sampling the voltage amplitude at the center point of the eye diagram, and use the bit error rate or eye opening as error information.

[0049] Eye opening is a parameter for evaluating the time-domain characteristics of signal quality, reflecting the timing jitter and amplitude distortion of the signal.

[0050] In some more specific examples, the bit error rate (BER) statistics window can be set to 1000 symbol periods to balance real-time performance with statistical accuracy; the eye diagram opening sampling point is at the center of the unit interval, and the voltage amplitude is calculated by comparing peak and trough values. The two indicators can be weighted and fused, for example, emphasizing BER in high-speed scenarios and emphasizing eye diagram opening in high-noise scenarios, with the weighting factor dynamically adjusted by the adaptive engine.

[0051] In these embodiments, the bit error rate (BER) serves as an indicator of signal quality and directly reflects the equalization effect. By monitoring the BER in real time, the adaptive engine can adjust parameters promptly to avoid signal degradation. The eye diagram opening directly represents signal integrity, enabling the adaptive engine to optimize equalization parameters to maximize the eye diagram opening. This feedback mechanism based on visual quality improves the reliability of high-speed signal transmission.

[0052] Optionally, in some embodiments of this application, the adaptive engine is further configured to: perform initial channel impulse response analysis when the error level represented by the feedback signal exceeds a preset threshold, identify the largest group region of non-peak regions in the channel impulse response through real-time online scanning, and determine the initial tap group position of the floating tap feedforward equalizer based on the identification result.

[0053] It should be noted that the preset threshold refers to the triggering condition set according to system requirements. When the performance index exceeds the threshold, a specific processing procedure is started. Channel impulse response analysis refers to the process of extracting channel characteristic parameters through algorithm processing.

[0054] In some more specific examples, the preset threshold can be set to a bit error rate of 1e-4 or an eye diagram opening decrease of 20%, and the trigger condition can be configured programmably; the initial channel analysis includes a full bandwidth scan, using a maximum likelihood estimation algorithm to locate the secondary peak region, and the analysis results are stored in the configuration register for use by the multiplexer. The entire process is completed within 100μs, ensuring rapid adaptation.

[0055] In these embodiments, the threshold-triggered initialization mechanism ensures the system's rapid adaptation to channel changes, which helps to solve the problem of cold start adaptation delay, enables floating taps to quickly locate key reflection points, and improves system robustness.

[0056] Optionally, in some embodiments of this application, the apparatus further includes: a microcontroller connected to the adaptive engine, configured to control the adaptive engine to perform channel impulse response analysis during the initialization phase to determine the initial tap group position of the floating tap feedforward equalizer.

[0057] It should be noted that a microcontroller (MCU) can be a chip that integrates a processor core, memory, and peripheral interfaces, and the initialization phase refers to the process of configuring basic parameters after the system starts up.

[0058] In some more specific examples, the microcontroller may employ an ARM Cortex-M series core, running dedicated firmware to manage the initialization process. The firmware implements channel scan scheduling, data caching, and result parsing, and interacts with the adaptive engine via an SPI interface. The microcontroller can also store historical channel data for prediction optimization, improving initialization accuracy.

[0059] In these embodiments, the introduction of the MCU enables fine control of the initialization process, and the accuracy of tap position optimization is ensured through firmware-managed initial analysis.

[0060] Reference Figure 2 As shown in the embodiments of this application, a serial signal receiving system is also provided. This SerDes architecture is an adaptive equalization processing system for high-speed serial communication / signal reception. Channel insertion loss / reflection causes inter-symbol interference (ISI). Through a full-link design of "analog front-end preprocessing → digital quantization and clock synchronization → hybrid equalization compensation → adaptive optimization," low bit error rate and high reliability digital signal recovery are achieved. Specifically, the system includes: An analog front end (AFE) is used to receive and preprocess analog signals. Specifically, the analog front end serves as the starting point for signal processing, preprocessing the input analog signals (such as differential electrical / optical signals in a high-speed serial link).

[0061] An analog-to-digital converter (ADC) is connected to the analog front end and is used to convert preprocessed analog signals into digital signals. (The corresponding part is in...) Figure 2 In the analog-to-digital converter section, a comparator can also be used. The comparator first performs an analog comparison on the analog signal output by the AFE. A comparator is used in a pure analog architecture. The analog-to-digital architecture converts the analog signal preprocessed by the AFE into a discrete-time digital sequence, supporting subsequent equalization and decision-making in the digital domain.

[0062] A clock and data recovery unit (CDR), connected to an analog-to-digital converter (ADC), is used to extract a clock signal from a digital signal and output a synchronized data signal. Specifically, the clock and data recovery unit extracts a synchronization clock from a digitized serial data stream and performs "data sampling timing" based on the clock.

[0063] The signal equalization device of the first aspect or any corresponding embodiment described above is connected to the clock data recovery module for equalizing the synchronized data signal.

[0064] The decision device (DFE) is connected to the signal equalizer to receive the equalized data signal and to the clock data recovery unit to receive the clock signal. It uses the clock signal to make decisions on the equalized data signal and outputs digital data. Specifically, the DFE performs hard / soft decision processing on the digital signal after FFE processing, mapping the "noisy, ambiguous sequence" to a "deterministic logic level" (such as "0 / 1" in NRZ code, and "0 / 1 / 2 / 3" in PAM4). Its output serves as both the recovered valid data output and the feedback to the DFE, acting as the input to the DFE.

[0065] The Decision Feedback Equalizer (DFE), connected to the Decision Unit, performs feedback equalization based on the decision results to eliminate inter-symbol interference (ISI). Specifically, the DFE utilizes the impact of the "decided result" on future symbols of the current symbol for feedback equalization, offsetting the long-tail effect of ISI (the problem of the current symbol being interfered with by subsequent symbols). The feedback coefficients of the DFE are adaptively adjusted by the Adaptation Engine, forming a hybrid architecture with the FFE of "feedforward compensation for general distortion + feedback to eliminate long-tail ISI," significantly improving the recovery capability of strong distortion channels.

[0066] More specifically, in the aforementioned signal equalization device, the Adaptation Engine is responsible for monitoring signal quality (such as eye diagram opening, error, etc.) and dynamically adjusting the equalization parameters. It receives feedback from the Decision Device's output and updates the filter coefficients / parameters of fixed-FFE, FT-FFE, and DFE in real time to achieve adaptive equalization parameters.

[0067] Both fixed-tap feedforward equalizers (FFEs) and floating-tap feedforward equalizers (FT-FFEs) are feedforward equalizers (FFEs), but they are designed to compensate for ISI at different locations. Fixed-FFE uses predefined fixed locations to pre-configure the equalization coefficient positions for "typical channels" (such as standard copper cables and backplanes), compensating for ISI at the peak position, minimizing ISI noise before the decision device, and reducing system complexity. FT-FFE, after initialization and adaptation by the MCU firmware, dynamically generates the ISI compensation positions for the filter coefficients based on the channel feature template (channel frequency / time domain features extracted in real time by the Adaptation Engine), specifically compensating for scenario-specific multipath interference (ISI) and improving the overall signal-to-noise ratio of the equalized signal.

[0068] In the above embodiments, the hybrid design of "fixed parameter fallback + feature template optimization" takes into account both "general scenario performance" and "adaptability to different scenarios", balancing hardware complexity and algorithm flexibility.

[0069] Reference Figure 3 As shown, the adaptive floating tap FFE in this application is composed of a fixed FFE and a floating FFE (hereinafter referred to as FT-FFE) stage. The fixed FFE compensates for the determined and regular ISI and processes the taps near the main curve of the channel. The tap position is fixed, which reduces the length compared with the fixed tap scheme of related technical solutions.

[0070] During the floating FFE phase, the secondary peak group of the channel is dynamically selected. The secondary peak group is the long trailing ISI of the original Main Cursor. For example, positions Px1~Px4 represent one region within a certain channel reflection location set, Py1~Py4 represent another secondary peak group region within the same channel reflection set, and so on. The tap positions can float horizontally, even to positions further than Py4. The floating taps can be grouped, such as 3 taps per group, for a total of 4 groups, etc. The floating taps can float according to the tap position.

[0071] In some embodiments of this application, the application utilizes hardware to calculate ISI and fixed TAP, performs firmware initialization to execute floating FT-FFE position, obtains maximum ISI position, and uses hardware to implement LMS algorithm to calculate FFE coefficients.

[0072] This application uses a fixed tap to cover the main peak region and a floating tap to cover the long-tailed secondary peak region, thus combining the stability of a fixed tap with the flexibility of a floating tap. The corresponding channel ISI diagram is shown in the reference diagram. Figure 3 As shown, a suitable peak region is selected as a fixed tap based on each connector (channel scenario).

[0073] The signal equalization device in this application employs the FT-FFE algorithm, which repeatedly converges to achieve stable ISI position. (Refer to...) Figure 3 The corresponding position in the diagram shows adaptive convergence across different channels, and stable convergence position within the same scene. The actual ISI position is referenced. Figure 4 The horizontal axis is shown in the graph. Different charts represent the FT-FFE convergence results of different SerDes channels under the same scenario. It can be seen that the horizontal axis is within a relatively fixed range under the same scenario. Figure 3 Two secondary peak regions in the middle; Reference Figure 4As shown, the taps of different channels converge within the same ISI (horizontal axis) interval, i.e., the same secondary peak region position, and the coefficient values ​​(vertical axis) are basically consistent, which illustrates the stability of the overall FT-FFE software and hardware algorithm calculation.

[0074] Reference Figure 5 As shown in the embodiments of this application, a signal equalization method embodiment is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0075] This embodiment provides a signal equalization method that can be used in the aforementioned signal equalization device. Figure 5 This is a flowchart of a signal equalization method according to an embodiment of this application, referred to... Figure 5 As shown, the process includes the following steps: Step S201: The input digital signal is equalized by the fixed-tap feedforward equalizer to compensate for inter-symbol interference in the main peak region of the channel impulse response. Step S202: The input digital signal is equalized by the floating tap feedforward equalizer to compensate for the inter-symbol interference in the secondary peak region of the channel impulse response. Specifically, step S202 includes: dynamically switching the activated tap group based on the time-domain characteristics of the channel impulse response using a multiplexer.

[0076] It should be noted that the dynamic switching process is a complete hardware and software collaborative adaptive process. (Refer to...) Figure 2 As shown, the system monitors the feedback signal from the decision maker in real time through an adaptive engine. When signal quality degradation is detected, the channel state analysis process is immediately triggered. This process first extracts the time-domain features of the channel impulse response to identify the secondary peak interference region outside the main peak region.

[0077] Reference Figure 3 As shown, the system establishes a dynamic mapping relationship between the identified secondary peak regions and floating tap groups. Each secondary peak region (such as Px1-Px4, Py1-Py4, etc.) corresponds to a specific tap group. This mapping relationship enables the system to accurately locate long-tailed ISIs in the channel. When channel conditions change, the adaptive engine re-analyzes the energy distribution and location characteristics of the secondary peak regions and generates corresponding control signals.

[0078] After receiving control signals from the adaptive engine, the multiplexer executes the specific switching operation. This process involves activating the tap group that best matches the current sub-peak region while disabling other unnecessary tap groups. Through this selective activation mechanism, the system can concentrate limited hardware resources on compensating for interference components that have the greatest impact on signal quality, thereby achieving on-demand resource allocation.

[0079] The entire dynamic switching process adopts a closed-loop control architecture, referring to... Figure 6 The hardware and software collaborative process is illustrated. During the initialization phase, the system establishes a basic mapping relationship through full-channel scanning. During normal operation, it continuously optimizes and updates the coefficients corresponding to the positions (adaptive convergence). The hardware is responsible for high-speed signal processing and rapid handover execution, while the software firmware undertakes policy decision-making and learning optimization. By analyzing historical channel data, it identifies the secondary peak regions of the current channel, further improving the accuracy and efficiency of handover. This design enables the system to not only respond quickly to instantaneous changes in the channel but also to optimize the handover strategy through long-term learning, thus maintaining excellent balanced performance in various complex application scenarios.

[0080] Actual test results refer to Figure 4 As shown, this dynamic switching mechanism exhibits good stability and adaptability under different channel conditions, maintaining a consistent convergence position within the same scenario, effectively improving the overall system performance. This dynamic switching mechanism, based on hardware and software collaboration, fundamentally solves the limitations of traditional fixed-tap FFE in dealing with complex and variable channel environments, achieving the optimal balance between resource efficiency and balanced performance.

[0081] Step S203: The digital signal after being equalized by the fixed tap feedforward equalizer and the floating tap feedforward equalizer is input to the decision unit to obtain a feedback signal. The adaptive engine generates error information characterizing the signal quality based on the reference benchmark and the feedback signal.

[0082] Specifically, in step S203 above, generating error information characterizing signal quality based on the feedback signal includes: Step S2031: Based on the feedback signal from the decision unit, the bit error rate is calculated by counting the number of erroneous bits per unit time, and / or the eye opening is calculated by sampling the voltage amplitude at the center point of the eye diagram; Step S2032: Use the calculated bit error rate or eye diagram opening as error information.

[0083] Step S204: Based on the error information, adjust the tap coefficients of the fixed tap feedforward equalizer and the tap coefficients of the activated tap groups in the floating tap feedforward equalizer through the adaptive engine.

[0084] Specifically, in step S204 above, the step of adjusting the tap coefficients based on the error information includes: Step S2041: The error information is processed by the least mean square algorithm to adjust the tap coefficients of the fixed tap feedforward equalizer and the floating tap feedforward equalizer.

[0085] Optionally, in some embodiments of this application, the signal equalization method further includes: Step a1: When the error level represented by the feedback signal exceeds a preset threshold, the largest group region of the non-peak region in the channel impulse response is identified by real-time online scanning to determine the initial tap group position of the floating tap feedforward equalizer.

[0086] Reference Figure 6 As shown, the specific calculation process for ISI and FT-FFE is as follows: It relies on software calculations (such as reading the channel status every 100ns → calculating coefficients → writing to the register), resulting in high response latency (1 microsecond). After initializing the tap position with software, the tap position is fixed, and the coefficients are adjusted by hardware.

[0087] The software FT-FFE algorithm for searching the maximum ISI is as follows: 1) Initialize the basic initialization operations of the FT-FFE module, including register reset, memory allocation, hardware state machine initialization, etc., to prepare the runtime environment for subsequent function execution.

[0088] 2) Tap selection Enable hardware circuitry to perform related calculations on the channel impulse response (IPR) of the entire link channel, providing data for tap selection. Based on the collected data, execute the optimal tap selection algorithm (such as based on the maximum likelihood criterion or the minimum mean square error criterion) to determine the activation positions of the floating FFE group (such as the specific coordinates of the fixed tap in the main peak area + the floating tap in the secondary peak area), and mark the taps to be activated.

[0089] 3) FT-FFE Selection Update Based on the tap selection result, the control signal of the multiplexer (MUX) is updated, and the delay line symbol corresponding to the selected tap is routed to the adder array, thus completing the final activation of the equalizer tap configuration.

[0090] The signal equalization method provided in this application can be executed based on the signal equalization device provided in any embodiment of this application, and has the corresponding beneficial effects of the device, which will not be described in detail here.

[0091] Although embodiments of this application have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of this application, and all such modifications and variations fall within the scope defined by the appended claims.

Claims

1. A signal equalization device, characterized in that, include: A fixed-tap feedforward equalizer is used to equalize the input digital signal to compensate for inter-symbol interference in the main peak region of the channel impulse response. A floating tap feedforward equalizer receives the input digital signal. The floating tap feedforward equalizer includes a multiplexer and multiple tap groups. Each tap group corresponds to a sub-peak region in the channel impulse response. The multiplexer is configured to dynamically switch the activated tap group based on the time-domain characteristics of the channel impulse response to compensate for inter-symbol interference in the sub-peak region. An adaptive engine is connected to both the fixed-tap feedforward equalizer and the floating-tap feedforward equalizer, and receives feedback signals from the decision unit. It is configured to generate error information characterizing signal quality based on a reference benchmark and the feedback signals, and adjust the tap coefficients of the fixed-tap feedforward equalizer and the tap coefficients of the activated tap groups in the floating-tap feedforward equalizer according to the error information.

2. The signal equalization device according to claim 1, characterized in that, Each tap group in the floating tap feedforward equalizer contains at least two taps, and the multiplexer is configured to uniformly activate or disable all taps within the same tap group.

3. The signal equalization device according to claim 1, characterized in that, The secondary peak region is the peak region in the channel impulse response whose amplitude is lower than the main peak but higher than the noise threshold.

4. The signal equalization device according to claim 1, characterized in that, The adaptive engine is configured to process the error information using a least mean square algorithm to adjust the tap coefficients of the fixed tap feedforward equalizer and the floating tap feedforward equalizer.

5. The signal equalization device according to claim 1, characterized in that, The adaptive engine is further configured to: when the error level represented by the feedback signal exceeds a preset threshold, perform initial channel impulse response analysis, identify the largest group region of non-peak regions in the channel impulse response through real-time online scanning, and determine the initial tap group position of the floating tap feedforward equalizer based on the identification result.

6. The signal equalization device according to claim 1, characterized in that, The device further includes: A microcontroller, connected to the adaptive engine, is configured to control the adaptive engine to perform channel impulse response analysis during the initialization phase to determine the initial tap group position of the floating tap feedforward equalizer.

7. A serial signal receiving system, characterized in that, include: Analog front end, used to receive analog signals and perform preprocessing; An analog-to-digital converter, connected to the analog front end, is used to convert the preprocessed analog signal into a digital signal; A clock data restorer, connected to the analog-to-digital converter, is used to extract the clock signal from the digital signal and output the synchronized data signal; The signal equalization device as described in any one of claims 1 to 6 is connected to the clock data recovery module and is used to equalize the synchronized data signal; A decision device, connected to the signal equalization device to receive the equalized data signal and connected to the clock data recovery device to receive the clock signal, for making a decision on the equalized data signal using the clock signal and outputting digital data; A decision feedback equalizer, connected to the decision device, for performing feedback equalization based on the decision result to eliminate post-cursor inter-symbol interference.

8. A signal equalization method, characterized in that, Comprising: Equalizing the input digital signal through the fixed-tap feed-forward equalizer to compensate for the inter-symbol interference in the main peak region of the channel impulse response; Equalizing the input digital signal through the floating-tap feed-forward equalizer to compensate for the inter-symbol interference in the secondary peak region of the channel impulse response; Inputting the digital signal equalized by the fixed-tap feed-forward equalizer and the floating-tap feed-forward equalizer into the decision device to obtain a feedback signal, and generating error information characterizing the signal quality by the adaptive engine based on a reference benchmark and the feedback signal; Adjusting the tap coefficients of the fixed-tap feed-forward equalizer and the tap coefficients of the activated tap groups in the floating-tap feed-forward equalizer by the adaptive engine according to the error information.

9. The signal equalization method according to claim 8, wherein The equalizing the input digital signal through the floating-tap feed-forward equalizer to compensate for the inter-symbol interference in the secondary peak region of the channel impulse response includes: Dynamically switching the activated tap groups through a multiplexer based on the time-domain characteristics of the channel impulse response.

10. The signal equalization method according to claim 8, wherein The step of adjusting the tap coefficients according to the error information includes: Processing the error information through the least mean square algorithm to adjust the tap coefficients of the fixed-tap feed-forward equalizer and the floating-tap feed-forward equalizer.

11. The signal equalization method according to claim 10, wherein The signal equalization method further includes: When the error level represented by the feedback signal exceeds a preset threshold, identifying the largest group region in the non-main peak region of the channel impulse response through real-time online scanning to determine the initial tap group position of the floating-tap feed-forward equalizer.

Images