Interference depth determination method and apparatus, and storage medium

Abstract

The application discloses a method and device for determining an intersymbol interference depth and a storage medium. The method comprises: performing transient simulation on a signal channel to obtain an impulse response; generating a candidate truncation point on a time axis based on the impulse response and determining characteristic indexes, including: cumulative energy determined based on the strength of the impulse response accumulated to the candidate truncation point, residual peak amplitude determined based on the maximum response amplitude after the candidate truncation point, and truncation error determined based on the difference between the cumulative energy up to the candidate truncation point and the overall energy; determining a judgment parameter of the characteristic indexes based on a target accuracy or system specification; taking the candidate truncation point meeting the judgment parameter as an effective memory length and a target intersymbol interference depth based on the characteristic indexes and the judgment parameter; establishing a time window corresponding to the target intersymbol interference depth, performing a statistical convolution operation, and outputting a statistical analysis result. The application has the technical effects of improving the accuracy and efficiency of intersymbol interference depth simulation.

Description

TECHNICAL FIELD

[0001] Embodiments of the present disclosure relate to the technical field of high-speed digital signal integrity simulation, and in particular to a method and device for determining inter-symbol interference depth and a storage medium. BACKGROUND

[0002] With the continuous improvement of the bandwidth of high-speed digital communication interfaces, signal integrity analysis has become an important part of high-speed circuit design. In such systems, due to effects such as channel loss, reflection and crosstalk, the waveform at the receiving end is often affected by the tail of the previous and next bits, forming inter-symbol interference (ISI), which leads to eye closure, increased jitter and rising bit error rate. The existing statistical eye diagram analysis method usually uses a fixed ISI depth or an empirically set memory length for simulation to truncate the channel impulse response. However, a fixed truncation length cannot take into account the differences in different channel characteristics: a value that is too small will ignore the far-end interference, resulting in insufficient accuracy, and a value that is too large will increase the computational overhead and reduce the simulation efficiency. Therefore, how to adaptively determine the appropriate inter-symbol interference depth while maintaining the accuracy of statistical analysis, to balance the simulation accuracy and computational efficiency, has become a technical problem to be solved in the field of high-speed digital signal integrity. SUMMARY

[0003] Therefore, embodiments of the present disclosure provide a method and device for determining inter-symbol interference depth and a storage medium to improve the simulation efficiency and accuracy of inter-symbol interference depth. In a first aspect, a method for determining inter-symbol interference depth is provided for statistical analysis of high-speed digital signals, comprising: performing transient simulation on a signal channel to obtain an impulse response; generating candidate truncation points on a time axis based on the impulse response, and determining a feature index for characterizing convergence, the feature index including: accumulated energy, residual peak amplitude, and truncation error, wherein the accumulated energy is determined based on the strength of the impulse response accumulated to the candidate truncation point, the residual peak amplitude is determined based on the maximum response amplitude after the candidate truncation point, and the truncation error is determined based on the difference between the accumulated energy up to the candidate truncation point and the overall energy; determining the determination parameters of the feature index based on the target accuracy or system specification, the determination parameters including: a first threshold for determining the accumulated energy, a second threshold for determining the residual peak amplitude, and a third threshold for determining the truncation error; based on the feature index and the determination parameters, the candidate truncation point that meets the determination parameters is taken as the effective memory length and as the target inter-symbol interference depth; establishing a time window corresponding to the target inter-symbol interference depth, performing statistical convolution operation, and outputting statistical analysis results, the statistical analysis results including: one or more of an eye diagram, a bit error rate, or a bathtub curve.

[0004] The above code interference depth determination method can accurately cover the response interval with actual interference contribution under different channel characteristics through the joint determination of the three types of indicators, i.e., cumulative energy proportion, residual peak amplitude and truncation error, thereby avoiding the over-idealization of the eye diagram or the underestimation of the bit error rate caused by premature truncation.

[0005] Optionally, the candidate truncation points are configured to be sequentially increased by a preset step along the time axis or the unit interval sequence.

[0006] Optionally, the determination parameters further include a maximum delay upper limit of the memory length, used to limit the search range of the candidate truncation points, and when there is no candidate truncation point satisfying the determination parameters in the maximum delay upper limit of the memory length, the candidate truncation point at the maximum delay upper limit of the memory length is used for truncation, and an accuracy limitation prompt is output.

[0007] Optionally, the candidate truncation point satisfying the determination parameters is used as the effective memory length, and the method further includes: in the multiple candidate truncation points satisfying the determination parameters, the first candidate truncation point satisfying the condition is used as the effective memory length.

[0008] Optionally, the code interference depth determination method further includes: determining the effective memory length based on the candidate truncation point satisfying the determination parameters and a safety margin coefficient, wherein the safety margin coefficient is determined based on a noise margin, a crosstalk level or an actual working condition.

[0009] Optionally, the code interference depth determination method further includes: after completing the statistical analysis based on the current target code interference depth, the first opening degree indicator of the eye diagram and the first estimation indicator of the bit error rate are obtained, and the second opening degree indicator and the second estimation indicator of the previous statistical analysis are compared by difference to determine the first difference value and the second difference value; when the first difference value is less than the eye diagram opening degree threshold and the second difference value is less than the bit error rate threshold, it is determined that the truncation point corresponding to the current target code interference depth converges.

[0010] Optionally, the time window corresponding to the target code interference depth is established, including: making the time window cover the signal edge sampling region and the subsequent effective tail interval.

[0011] Optionally, the method further includes: when the signal channel has a multi-edge sampling characteristic, the candidate truncation points of the rising edge impulse response and the falling edge impulse response are determined respectively, and the maximum candidate truncation point is used as the effective memory length of the signal channel.

[0012] In a second aspect, a code interference depth determination apparatus is provided for statistical analysis of high-speed digital signals, comprising: a simulation unit configured to perform transient simulation on a signal channel to obtain an impulse response; a determination unit configured to generate candidate cut-off points on a time axis based on the impulse response, and determine a feature index for characterizing convergence, the feature index comprising: accumulated energy determined based on strength of the impulse response accumulated to the candidate cut-off points, residual peak amplitude determined based on maximum response amplitude after the candidate cut-off points, and cut-off error determined based on difference between the accumulated energy up to the candidate cut-off points and overall energy; a judgment unit configured to determine a judgment parameter of the feature index based on a target accuracy or system specification, the judgment parameter comprising: a first threshold for determining the accumulated energy, a second threshold for determining the residual peak amplitude, and a third threshold for determining the cut-off error; and configured to determine, as an effective memory length and as a target code interference depth, the candidate cut-off point satisfying the judgment parameter based on the feature index and the judgment parameter; and an operation unit configured to establish a time window corresponding to the target code interference depth, perform statistical convolution operation, and output a statistical analysis result, the statistical analysis result comprising: one or more of an eye diagram, a bit error rate, or a bathtub curve.

[0013] In a third aspect, a computer-readable storage medium is provided, comprising a memory having instructions stored thereon, the instructions being configured to be read by a processor to implement the code interference depth determination method provided in the first aspect. BRIEF DESCRIPTION OF DRAWINGS

[0014] The accompanying drawings for use in the description of embodiments of the present disclosure are briefly described as follows:

[0015] Figure 1 A flowchart of a code interference depth determination method provided in some embodiments of the present application is shown;

[0016] Figure 2 A flowchart of other steps of a code interference depth determination method provided in some embodiments of the present application is shown;

[0017] Figure 3 An eye diagram comparison under different code interference depth conditions provided in some embodiments of the present application is shown;

[0018] Figure 4 A structural diagram of a code interference depth determination apparatus provided in some embodiments of the present application is shown. DETAILED DESCRIPTION

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the following will describe the embodiments of the present disclosure with reference to the drawings. The drawings in the following description are only some embodiments of the present disclosure, and for those skilled in the art, other drawings can be obtained from these drawings without creative labor, and other embodiments can be obtained, and the adjustments and improvements made without departing from the concept of the present disclosure are within the protection scope of the present disclosure.

[0020] In order to make the drawings simple, each drawing only schematically shows the part related to the embodiments, and it does not represent the actual structure of the product. In addition, in order to make the drawings simple and easy to understand, in some drawings, only some parts with the same structure or function are schematically shown, and there can be more or less parts with the same structure or function.

[0021] In the present disclosure, unless otherwise explicitly specified and limited, ordinal words such as "first", "second", etc. are only used to distinguish the description of the associated objects, and cannot be understood as indicating or implying the relative importance or order between the associated objects; in addition, it also does not represent the number of the associated objects. "Multiple" includes two or more, and other quantifiers are similar. " / " is used to describe the relationship between the associated objects, which represents the "or" relationship between the associated objects. "And / or" is used to describe the relationship between the associated objects, which includes any combination relationship between the associated objects, for example, "a and / or b" includes: "a alone", "b alone", or "a and b". "One or more" or "at least one" of multiple objects means any object or any combination of multiple objects, for example, "one or more of a1, a2, a3" or "at least one of a1, a2, a3" includes: "a1 alone", "a2 alone", "a3 alone", "a1 and a2", "a1 and a3", "a2 and a3", or "a1, a2 and a3".

[0022] With the continuous increase of bandwidth of high-speed digital communication interface, signal integrity analysis has become an important part of high-speed circuit design. Due to channel loss, reflection, crosstalk and imperfect terminal matching, etc., the input signal will appear trailing and echo during transmission, which makes the front and back bits superimpose each other, forming inter-symbol interference (ISI), and leading to eye closure, jitter increase and bit error rate rise. In high-speed interface design and verification, statistical eye diagram analysis is a widely used signal integrity evaluation method. Based on the impulse response or step response of the channel, the method predicts the signal waveform distribution under any bit sequence through statistical convolution to estimate the eye opening and bit error rate (BER). In the simulation process of existing DDR and high-speed serial bus, statistical eye diagram analysis usually uses fixed or default ISI depth to determine the effective memory length of impulse response, that is, the channel response is simulated within a fixed time window, and the ISI contribution outside the window is assumed to be negligible. By default, eye diagram simulation often sets a default time window within a certain tolerance range according to experience or standard template. For example, the impulse response of the channel generally decays to a negligible level after about 8 unit intervals (UI), and if each UI contains 64 sampling points, the default only considers 8×64=512 sampling points. Although this truncation method can control the amount of calculation, it may not match the simulation accuracy and efficiency required by the actual channel. In the actual simulation process, users can only adjust the ISI depth by experience. If the set value is too small, the residual ISI contribution of the far end of the channel will be ignored, which will lead to the important tail of the impulse response not being considered, and thus underestimate the risk of eye closure. For example, in a channel with long delay echo or slow energy decay, premature truncation will miss those that only have a very low bit error rate (such as 1×10 -16) interference factors that only appear under certain conditions, making the simulation results too idealized. If you want to compensate for this error through brute force time domain simulation, the amount of calculation is huge and unfeasible. On the contrary, if the ISI depth value is too large, the system will perform redundant calculations on a large number of stable impulse responses, significantly increasing the calculation time and resource consumption, while the convergence accuracy is not significantly improved. Because the energy contribution of the impulse response after reaching the steady state is close to zero, the impact on the subsequent results can be ignored. Therefore, the traditional method is difficult to balance between accuracy and efficiency, and the parameter setting has high uncertainty, which needs to be repeatedly calculated or guessed by humans. In summary, the existing technology mainly relies on fixed or empirically set ISI depth, lacks adaptive control mechanism based on channel characteristics, and makes it difficult to balance the accuracy and efficiency of statistical eye diagram analysis. When the set value is too small, the interference is underestimated, and when it is too large, the simulation time and calculation resources are wasted. Therefore, the present application proposes a code interference depth determination method, device, and storage medium, which determines the effective memory length of code interference by comprehensively determining the energy accumulation, residual peak amplitude, and truncation error, etc. characteristics, avoiding redundant calculation and improving simulation convergence and reliability.

[0023] The following will be described with reference to the accompanying drawings:

[0024] Figure 1 A flowchart of a code interference depth determination method provided in some embodiments of the present application is shown. The code interference depth determination method is used for statistical analysis of high-speed digital signals, and at least includes the following steps:

[0025] S110: performing transient simulation on the signal channel to obtain the impulse response;

[0026] S120: based on the impulse response, generating a candidate truncation point on the time axis, and determining a feature index for characterizing convergence, the feature index including: accumulated energy, residual peak amplitude, and truncation error, wherein the accumulated energy is determined based on the strength of the impulse response accumulated to the candidate truncation point, the residual peak amplitude is determined based on the maximum response amplitude after the candidate truncation point, and the truncation error is determined based on the difference between the accumulated energy up to the candidate truncation point and the overall energy;

[0027] S130: determining the determination parameters of the feature index based on the target accuracy or system specification, the determination parameters including: a first threshold for determining the accumulated energy, a second threshold for determining the residual peak amplitude, and a third threshold for determining the truncation error;

[0028] S140: based on the feature index and the determination parameters, taking the candidate truncation point that meets the determination parameters as the effective memory length, and as the target code interference depth;

[0029] S150: Establish a time window corresponding to the target intersymbol interference depth, perform statistical convolution operation, and output statistical analysis results, including one or more of eye diagram, bit error rate, or bathtub curve.

[0030] In the above embodiment of the intersymbol interference depth determination method, first, the transient simulation is performed on the target signal channel, the standard excitation is applied to characterize the channel characteristics, and the impulse response of the output of the receiving end at the sampling resolution is recorded. The impulse response can include waveform samples during the rising and falling edges, and the decay segment before reaching the steady state. On the time axis corresponding to the impulse response sequence, candidate truncation points can be sequentially generated according to the sampling resolution or the UI granularity. For each candidate truncation point, three types of feature indicators characterizing convergence can be used for calculation. The cumulative energy refers to the energy proportion up to the point obtained by gradually accumulating and converting the response intensity before the candidate truncation point as a boundary. For example, taking the signal channel of the target link as the object, applying the excitation and starting the transient simulation, obtaining the impulse response of the channel output as h(n), when determining the cumulative energy E i (N), formula 1 can be referred to:

[0031]

[0032] Where i represents the current candidate truncation point position, and n represents the truncation point in the impulse response.

[0033] The residual peak amplitude h res (N) refers to the maximum value in the amplitude absolute value extracted from the response samples after the candidate truncation point, which is used to characterize the tail residual intensity. When determining the residual peak amplitude h res (N), formula 2 can be referred to:

[0034]

[0035] The truncation error refers to the difference between the accumulated energy up to the candidate truncation point and the overall energy based on the overall energy of the response, which is used to characterize the proportion of energy missing caused by truncation. Therefore, when determining the truncation error, formula 3 can be referred to:

[0036]

[0037] The three types of indicators can be stored in scalar form for subsequent determination. When determining the determination parameters, threshold values corresponding to the three types of feature indicators can be configured according to the target accuracy or system specifications. The target accuracy is based on the result fineness and reliability required by statistical analysis, which can usually be determined by engineering targets, such as target bit error rate level, eye opening, minimum margin of width, allowable fluctuation of index change (convergence threshold), etc. The system specification refers to the external constraints of the applicable interface and test, such as interface type (DDR / PCIe / SerDes, etc.), data rate and UI setting, sampling resolution, maximum delay range allowed, equalization / determination mode, etc., which limit the analysis boundary and qualification criteria. Therefore, the first threshold value for cumulative energy ratio, the second threshold value for residual peak amplitude, and the third threshold value for truncation error in the determination parameters of the feature indicators can be reasonably determined by those skilled in the art based on the above two contents. For the candidate truncation points generated in sequence, the corresponding three types of feature indicators can be read one by one and compared with the determination parameters. For example, when a candidate truncation point that meets the first threshold value, the second threshold value, and the third threshold value appears, the candidate truncation point can be determined as the effective memory length and as the target code interval interference depth. With the signal amplitude V pp , the truncation point N is adjusted step by step to determine the candidate truncation point N that meets the determination parameters cand , as shown in formula 4:

[0038]

[0039] Based on the target code interval interference depth, a statistical time window corresponding thereto is established, so that the window covers the sampling region at the signal edge and the subsequent effective tail interval. Statistical convolution operation is performed in the time window to output statistical analysis results, which at least include one or more of eye diagram, bit error rate, and / or bathtub curve. To ensure traceability of the results, the target code interval interference depth, time window boundary, and corresponding determination parameters used for calculation can also be recorded by the system. Candidate point step adaptation: when the operation load is high or the response is close to steady state, the step of the candidate truncation point can be adaptively increased to shorten the search time; in the section where the response changes dramatically, the step remains consistent with the sampling resolution to avoid missed judgment. The embodiments of the present application can accurately cover the response interval with actual interference contribution under different channel characteristics through the joint determination of the three types of indicators: cumulative energy ratio, residual peak amplitude, and truncation error, thereby avoiding the idealization of the eye diagram or the underestimation of the bit error rate caused by premature truncation.

[0040] In some embodiments of the present application, the candidate truncation points are configured to be sequentially increased by a preset step along the time axis or unit interval sequence.

[0041] Unit Interval (UI) can be used to evaluate the energy convergence of the channel impulse response step by step and determine the appropriate truncation position. After the system completes the transient simulation of the signal channel and obtains the impulse response sequence, the system discretizes the impulse response according to the sampling time, and each discrete point can correspond to a sampling time or a unit interval position. Then, the system sets an initial step value according to the sampling resolution, and selects a plurality of sampling positions as candidate truncation points on the time axis with the step. The step value can be a fixed value, or it can be adaptively adjusted according to the speed of response change. When the impulse response changes dramatically or the energy accumulation curve grows rapidly, a smaller step (such as the same as the sampling resolution) can be set to ensure the calculation accuracy; when the response tends to be stable or decays slowly, the step can be appropriately increased to reduce the number of candidate points and the amount of calculation. For example, the initial value of the step can be set to an integer multiple of the number of sampling points in a unit interval (UI), for example, 16 or 32 sampling points per interval to generate a candidate truncation point. When the system traverses the candidate points, it calculates three types of feature indexes, including cumulative energy proportion, residual peak amplitude and truncation error for each truncation point, and compares them with the corresponding threshold. When the first candidate truncation point that meets the threshold requirements appears, it is determined that the position is the effective memory length.

[0042] In some embodiments of the present application, the determination parameters further include a maximum delay upper limit of the memory length, which is used to limit the search range of the candidate truncation point. When there is no candidate truncation point that meets the determination parameters in the maximum delay upper limit of the memory length, the candidate truncation point at the maximum delay upper limit of the memory length is used for truncation, and an accuracy limited prompt is output.

[0043] The maximum delay upper limit of the memory length represents the maximum response interval that can be searched on the time axis by the method of the embodiments of the present application, that is, the maximum value of the candidate truncation point in the method implementation process cannot exceed the upper limit. For example, when the formula 4 is used for determination, if N reaches the delay upper limit T max , the truncation at T max , and a warning of accuracy loss, or dynamically improve T maxThis parameter can be automatically determined based on interface specifications, signal rate, or bit error rate targets, or it can be manually configured by the user according to channel length and simulation accuracy requirements. For example, for a typical DDR interface or high-speed serial channel, the maximum delay limit can be set to 4 to 8 unit intervals (UI) to cover the main effective tailing interval. When performing incremental judgment of candidate truncation points, the system sequentially selects candidate truncation points from the starting point of the impulse response and calculates characteristic indicators for each candidate point (such as the cumulative energy percentage, residual peak amplitude, and truncation error provided in the above embodiment), and compares the results with the corresponding thresholds. When the first candidate truncation point that simultaneously satisfies all the judgment parameters appears, it is determined as the effective memory length. If no candidate truncation point that satisfies all the conditions appears when traversing to the maximum delay limit, the system automatically truncates at the maximum delay limit and takes the candidate truncation point corresponding to that position as the final target inter-symbol interference depth. To maintain the traceability of analysis results, the system can generate a precision limitation warning in the logs or reports when performing this operation. This warning explicitly indicates that the calculation did not fully meet the preset threshold before reaching the maximum delay limit, reminding the user to reconfigure the threshold or delay limit according to actual needs to improve accuracy. The introduction of the maximum delay limit effectively prevents the candidate cutoff point search process from becoming too long or failing to terminate when there are long delay echoes or slow energy decay in the channel. It also provides a unified upper limit constraint for different systems or specifications, ensuring the stability of inter-symbol interference depth determination and the controllability of computational resources.

[0044] In some embodiments of this application, the candidate cutoff point that satisfies the determination parameters is used as the effective memory length, and the method further includes: among a plurality of candidate cutoff points that satisfy the determination parameters, the first candidate cutoff point that satisfies the conditions is used as the effective memory length.

[0045] In some embodiments of this application, the method for determining inter-symbol interference depth further includes: determining the effective memory length based on candidate cutoff points that satisfy the judgment parameters and a safety margin coefficient, wherein the safety margin coefficient is determined based on noise margin, crosstalk level, or actual operating conditions.

[0046] In the above embodiments, when the candidate truncation point satisfying the cumulative energy, residual peak amplitude and truncation error determination conditions is determined on the time axis, the effective memory length is obtained. Since the actual signal may be affected by random noise, adjacent channel crosstalk or working condition changes (such as temperature, voltage, manufacturing deviation, etc.) during transmission, the effective memory length may be slightly lower than the true interference depth required under extreme conditions. Therefore, a safety margin coefficient can be introduced to adjust the effective memory length. For example, the safety margin coefficient can be a fixed proportion value, or it can be adaptively determined according to the current channel environment: when a higher noise margin or a significant crosstalk level is detected, a larger safety margin coefficient can be taken to ensure that more potential interference components are included; when the channel is in a stable or low-noise state, a smaller safety margin coefficient can be taken to reduce the calculation redundancy. For example, the safety margin coefficient can take a value between 1.05 and 1.20. The preliminary effective memory length is multiplied by the safety margin coefficient to obtain the final target intersymbol interference depth N opt , referring to formula 5:

[0047]

[0048] The intersymbol interference depth N opt can be used to establish a time window for statistical analysis, ensuring that the simulation range sufficiently covers the residual response below the noise boundary. In this way, even in the presence of actual uncertainties such as random noise, coupled interference or temperature drift, the statistical eye diagram analysis result can still maintain accuracy and consistency, avoiding misjudgment or error accumulation caused by premature truncation. This comprehensive strategy not only preserves the overall energy contribution but also considers the instantaneous peak interference and statistical error influence, thereby more robustly determining the intersymbol interference depth N opt . The intersymbol interference depth N opt can take the smallest N that satisfies the threshold condition, so that further increasing the intersymbol interference memory length has little effect.

[0049] In some embodiments of the present application, Figure 2 a flowchart showing other steps of an intersymbol interference depth determination method provided in some embodiments of the present application is shown. It also includes:

[0050] S210: After completing the statistical analysis based on the current target intersymbol interference depth, the first opening degree indicator of the eye diagram and the first error rate estimation indicator are obtained, and the second opening degree indicator and the second estimation indicator of the previous statistical analysis are compared, to determine the first difference and the second difference;

[0051] S220: When the first difference is less than the eye diagram opening degree threshold, and the second difference is less than the error rate threshold, it is determined that the truncation point corresponding to the current target intersymbol interference depth converges.

[0052] The eye opening index and the bit error rate estimation index obtained by the current round of statistical analysis are extracted respectively, and are denoted as a first eye opening index and a first estimation index P N respectively. Meanwhile, a second eye opening index and a second estimation index P N-1 obtained by the previous round of statistical analysis are read. The system calculates the difference between the two sets of results to obtain a first difference value for representing the change in the eye opening index in the consecutive two rounds of calculation, and a second difference value for representing the change in the bit error rate estimation in the consecutive two rounds of calculation. The difference value can be expressed in the form of absolute value or percentage to reflect the fluctuation amplitude of the statistical result. Referring to formula 6:

[0053]

[0054] Further, the first difference value is compared with a preset eye opening index threshold E eye , and the second difference value is compared with a preset bit error rate threshold E BER . When the first difference value is less than the eye opening index threshold E eye and the second difference value is less than the bit error rate threshold E BER , it is determined that the convergence of the corresponding truncation point of the target intersymbol interference depth has been reached. If any difference value exceeds the threshold, it can be considered that the statistical analysis has not yet converged, and the system continues to increment the candidate truncation point and repeats the statistical analysis until the convergence condition is met or the maximum delay upper limit of the memory length is reached. The eye opening index threshold and the bit error rate threshold can be obtained according to the target accuracy or system specification configuration. For example, in the high-precision analysis mode, the eye opening index threshold can be taken as below 0.5%, and the bit error rate threshold can be taken as below 1%; in the standard or fast mode, the thresholds can be correspondingly relaxed to balance the accuracy and calculation efficiency.

[0055] In some embodiments of the present application, the time window corresponding to the target intersymbol interference depth is established, including: covering the signal edge sampling region and the subsequent effective tail interval.

[0056] The intersymbol interference depth determination method further includes: when the signal channel has a multi-edge sampling characteristic, determining the candidate truncation points of the rising edge impulse response and the falling edge impulse response respectively, and taking the maximum candidate truncation point as the effective memory length of the signal channel.

[0057] The intersymbol interference depth N optAfterwards, a corresponding statistical analysis window can be established on the time axis according to the depth. The starting position of the time window can be aligned with the signal sampling time or the bit decision point, and the ending position extends to a time point after the target intersymbol interference depth. Such a design can ensure that the sampled data contains both the transient response at the signal flip and the tail waveform gradually decaying over time. In actual implementation, the length of the time window can be determined by the target intersymbol interference depth, and the total length of the time window is the time length or unit interval length corresponding to the depth. The system can perform statistical convolution operation in the window to generate statistical analysis results such as eye diagram, bit error rate and bathtub curve. For interfaces with double-edge triggering (such as DDR, EDGE or double data rate link), because the pulse response form and decay rate of the rising edge and the falling edge can be different, the rising edge pulse response and the falling edge pulse response can be extracted respectively, and the corresponding candidate cutoff points can be calculated independently. The calculation process is consistent with the single-edge example given in the foregoing embodiment, and also includes steps such as feature index calculation, threshold determination and convergence verification. When both sets of calculations are completed, the lengths of the candidate cutoff points of the rising edge and the falling edge can be compared, and the maximum value of the two is selected as the effective memory length of the signal channel as a whole. In this way, under any edge triggering condition, the statistical analysis covers the longest valid interference interval, preventing partial responses from being truncated and affecting the accuracy of the analysis results.

[0058] In some embodiments of the present application, the intersymbol interference depth determination method further includes a result checking and iteration process for verifying the reasonableness of the intersymbol interference depth determined by the adaptive algorithm and optimizing and adjusting the parameters according to the output results. After completing the statistical analysis and outputting the results such as eye diagram, bit error rate and bathtub curve, the system can perform comparative analysis on the key indicators of the eye diagram under different intersymbol interference depths to confirm whether the calculation results of the adaptive algorithm meet the expectations. For example, the comparative analysis can include: comparing the differences in eye opening, closing degree and noise distribution of the eye diagrams generated under the conditions of default depth, manually set depth and adaptive depth; verifying whether the eye diagram curve converges smoothly and the distortion degree is reduced under the intersymbol interference depth determined by the adaptive algorithm; and comparing whether the bit error rate distribution or bit error probability density is consistent with the channel characteristics and the target bit error rate. In some embodiments of the present application, the system can also output the decision process information in the simulation log or report, including data such as threshold judgment, cutoff point position, energy accumulation ratio, residual peak amplitude and truncation error, to improve the transparency of the algorithm and facilitate users to further adjust the parameters according to the output information. For example, Figure 3The eye diagram contrast conditions under different intersymbol interference depths provided in some embodiments of the present application are shown. From top to bottom, the simulation results correspond to the default depth, the manually set depth and the adaptive depth respectively. It can be clearly seen that in the adaptive case, the degree of eye diagram distortion is smaller, the curvature is smoother, and the opening degree is more matched with the actual noise characteristics of the channel, thereby more truly reflecting the comprehensive influence of noise and crosstalk in high-speed digital signal integrity analysis. When the contrast results show that the eye diagram curve obtained by the adaptive depth is smooth and the bit error rate distribution is stable, the system determines that the algorithm result is reasonable. If it is found that there is an abnormal convergence or a large deviation, the judgment threshold or safety margin can be automatically adjusted and the analysis can be re-executed according to the log information, so as to realize the self-correction and parameter iteration of the result.

[0059] Figure 4 A structural schematic diagram of an intersymbol interference depth determination device provided in some embodiments of the present application is shown. The intersymbol interference depth determination device is used for high-speed digital signal statistical analysis, and includes: a simulation unit 410, configured to perform transient simulation on a signal channel to obtain an impulse response; a determination unit 420, configured to generate a candidate truncation point on a time axis based on the impulse response, and determine a feature index used for representing convergence, the feature index including: accumulated energy, residual peak amplitude and truncation error, wherein the accumulated energy is determined based on the strength of the impulse response accumulated to the candidate truncation point, the residual peak amplitude is determined based on the maximum response amplitude after the candidate truncation point, and the truncation error is determined based on the difference between the accumulated energy up to the candidate truncation point and the overall energy; a judgment unit 430, configured to determine a judgment parameter of the feature index based on a target accuracy or a system specification, the judgment parameter including: a first threshold for determining the accumulated energy, a second threshold for determining the residual peak amplitude, and a third threshold for determining the truncation error; and configured to take the candidate truncation point meeting the judgment parameter as an effective memory length and as a target intersymbol interference depth based on the feature index and the judgment parameter; and an operation unit 440, configured to establish a time window corresponding to the target intersymbol interference depth, perform statistical convolution operation, and output a statistical analysis result, the statistical analysis result including: one or more of an eye diagram, a bit error rate or a bathtub curve.

[0060] Based on the same technical concept, the present application also provides a computer readable storage medium, including a memory, and instructions stored on the memory, the instructions being configured to be read by a processor to implement the intersymbol interference depth determination method provided in the above embodiments.

[0061] In the above embodiments, the description of each embodiment has its own focus, and the parts not described or recorded in detail in a certain embodiment can be referred to the related description of other embodiments. In addition, the above embodiments can be freely combined as needed.

Claims

1. A method for determining the depth of inter-symbol interference, characterized in that, Used for statistical analysis of high-speed digital signals, including: Perform transient simulation on the signal channel to obtain the impulse response; Based on the impulse response, candidate cutoff points are generated on the time axis, and characteristic indicators for summation convergence are determined. The characteristic indicators include: cumulative energy, residual peak amplitude, and truncation error. The cumulative energy is determined by summing the intensity of the impulse response up to the candidate cutoff point. The residual peak amplitude is determined by the maximum response amplitude after the candidate cutoff point. The truncation error is determined by the difference between the cumulative energy up to the candidate cutoff point and the overall energy. The determination parameters for the characteristic indicators are determined based on the target accuracy or system specifications. The determination parameters include: determining a first threshold for the cumulative energy, determining a second threshold for the residual peak amplitude, and determining a third threshold for the truncation error. Based on the feature index and the determination parameter, the candidate cutoff point that satisfies the determination parameter is taken as the effective memory length and as the target inter-code interference depth. Establish a time window corresponding to the target inter-code interference depth, perform statistical convolution operations, and output statistical analysis results, which include one or more of the following: eye diagram, bit error rate, or bathtub curve.

2. The method for determining the inter-symbol interference depth according to claim 1, characterized in that, The candidate cutoff points are configured to increase sequentially along the time axis or unit interval sequence by a preset step size.

3. The method for determining the inter-symbol interference depth according to claim 2, characterized in that, The determination parameters also include: the maximum delay limit of the memory length, which is used to limit the search range of the candidate cutoff point. When there is no candidate cutoff point that satisfies the determination parameters in the maximum delay limit of the memory length, the candidate cutoff point at the maximum delay limit of the memory length is used for truncation, and a precision limitation prompt is output.

4. The method for determining inter-symbol interference depth according to claim 3, characterized in that, The step of using the candidate cutoff point that satisfies the determination parameter as the effective memory length further includes: among the multiple candidate cutoff points that satisfy the determination parameter, using the first candidate cutoff point that satisfies the condition as the effective memory length.

5. The method for determining inter-symbol interference depth according to claim 4, characterized in that, Also includes: The effective memory length is determined based on the candidate cutoff point and safety margin coefficient that satisfy the determination parameters, wherein the safety margin coefficient is determined based on noise margin, crosstalk level or actual operating conditions.

6. The method for determining the inter-symbol interference depth according to any one of claims 1 to 5, characterized in that, Also includes: After completing the statistical analysis based on the current target inter-symbol interference depth, the first opening index of the eye diagram and the first estimated index of the bit error rate are obtained respectively, and the difference is compared with the second opening index and the second estimated index of the previous statistical analysis to determine the first difference and the second difference. When the first difference is less than the eye diagram opening threshold and the second difference is less than the bit error rate threshold, the truncation point corresponding to the current target inter-symbol interference depth is determined to converge.

7. The method for determining the inter-symbol interference depth according to any one of claims 1 to 5, characterized in that, The establishment of a time window corresponding to the target inter-code interference depth includes: making the time window cover the signal edge sampling area and its subsequent effective tailing interval.

8. The method for determining the inter-symbol interference depth according to any one of claims 1 to 5, characterized in that, Also includes: When the signal channel has multi-edge sampling characteristics, candidate cutoff points for rising edge pulse response and falling edge pulse response are determined respectively, and the largest candidate cutoff point is taken as the effective memory length of the signal channel.

9. A device for determining the depth of inter-symbol interference, characterized in that, Used for statistical analysis of high-speed digital signals, including: The simulation unit is used to perform transient simulation of the signal channel and obtain the impulse response; A determining unit is configured to generate candidate cutoff points on the time axis based on the impulse response, and determine characteristic indicators for summarizing convergence, the characteristic indicators including: cumulative energy, residual peak amplitude, and truncation error, wherein the cumulative energy is determined based on the intensity of the impulse response accumulated to the candidate cutoff point, the residual peak amplitude is determined based on the maximum response amplitude after the candidate cutoff point, and the truncation error is determined based on the difference between the cumulative energy up to the candidate cutoff point and the overall energy; The determination unit is used to determine the determination parameters of the feature index based on the target accuracy or system specifications. The determination parameters include: determining a first threshold for the cumulative energy, determining a second threshold for the residual peak amplitude, and determining a third threshold for the truncation error; and is used to, based on the feature index and the determination parameters, take the candidate truncation point that satisfies the determination parameters as the effective memory length and as the target inter-symbol interference depth. The computation unit is used to establish a time window corresponding to the target inter-code interference depth, perform statistical convolution operations, and output statistical analysis results, which include one or more of the following: eye diagram, bit error rate, or bathtub curve.

10. A computer-readable storage medium, characterized in that, The method includes a memory having instructions stored thereon, the instructions being configured to, when a processor reads the memory, implement the inter-symbol interference depth determination method according to any one of claims 1 to 8.

Images