CN119324768A - Hilbert transform-based time interval error jitter extraction method - Google Patents

本发明公开了基于希尔伯特变换的时间间隔误差抖动提取方法，涉及高速数据信号处理领域，当获取被测数据信号后，首先通过电平直方图统计，并确定相应的电平判决基准；其次对被测数据信号进行希尔伯特变换，获取被测数据信号跳变沿的时间信息；采用“粗频估计+精频估计”的方法提取被测数据信号准确的码元速率，进行高精度常频时钟恢复；依据上述恢复时钟进行高精度时间间隔误差提取，本发明提供的基于希尔伯特变换的时间间隔误差抖动提取方法，为高精度时间间隔误差抖动提供基础，准确的对被测数据信号进行抖动测量。

The invention discloses a time interval error jitter extraction method based on Hilbert transform, and relates to the field of high-speed data signal processing. After obtaining the measured data signal, firstly, a level histogram is used for statistics, and a corresponding level judgment reference is determined; secondly, the measured data signal is subjected to Hilbert transform to obtain the time information of the transition edge of the measured data signal; a "coarse frequency estimation + fine frequency estimation" method is used to extract the accurate symbol rate of the measured data signal, and a high-precision constant frequency clock is recovered; and high-precision time interval error is extracted according to the above-mentioned recovered clock. The time interval error jitter extraction method based on Hilbert transform provided by the invention provides a basis for high-precision time interval error jitter, and accurately performs jitter measurement on the measured data signal.

Hilbert transform-based time interval error jitter extraction method

Technical Field

The invention relates to the field of high-speed data signal processing, in particular to a time interval error jitter extraction method based on Hilbert transformation.

Background

The high-speed serial data link SerDes system is comprised of a transmitting end, a channel, and a receiving end, wherein the channel may comprise any physical transmission medium including copper wire, coaxial cable, fiber optic, and other mediums.

Jitter is a core index for measuring stability and reliability of data transmission, plays an important role in test analysis of ultra-large-scale integrated circuits, high-speed serial links and cloud computing, and time interval error TIE (TimeIntervalError) is initially defined in a g.810 recommendation issued by telecommunication standardization sector of the international telecommunication union and is widely used for testing instruments in the follow-up, and TIE is used as a basic unit for analyzing data jitter, and in a serial data system, TIE jitter is a phase difference between edges of a data signal and edges of a clock signal.

Since the high-speed serial data signal generally adopts an embedded clock transmission mode, after the receiving end receives the data signal, it is necessary to extract a clock signal embedded in the transmission data, and perform signal processing such as subsequent data recovery and jitter decomposition measurement on the data signal according to the extracted clock signal. In general, there are two methods of clock recovery at the receiving end, namely a normal frequency clock recovery method and a clock recovery method based on a phase-locked loop.

The clock recovery method based on the phase-locked loop can follow or track the low-frequency jitter transformation in the input signal, and in general, in order to analyze all jitter components such as random jitter, data correlation jitter and the like in the data signal, the extraction of the TIE is generally performed by adopting a normal-frequency clock recovery method, so that the extraction of an accurate time interval error TIE from the data signal is a precondition of a jitter analysis measurement technology, and the accuracy of the normal-frequency clock recovery also determines the accuracy of a jitter analysis measurement result. The high-precision TIE extraction is an important link for accurately evaluating the system performance, positioning jitter and improving the system design.

Published application number 202011468743.3: the PAM4 signal clock data recovery method based on the software phase-locked loop is based on an edge crossing method, clock recovery based on the phase-locked loop is realized by extracting a synchronous clock from a data signal, and in general, the TIE extracted based on the clock recovery of the phase-locked loop tracks low-frequency jitter of the data signal, namely ignores the low-frequency jitter in the data signal, and cannot extract all jitter information contained in the data signal, particularly when the data signal has serious intersymbol interference, crosstalk and other influences, the clock frequency obtained based on the edge crossing method has larger deviation.

The application number 201610287412.7 discloses "improved clock recovery for data signals", which is based on the technique of spectral line method, by deriving the data signal to determine edge information, a band-pass filter with a flat passband is applied to the data signal with SSC spread spectrum clock modulation, thereby improving the eye pattern of the data signal after the recovered clock and increasing the opening of the eye pattern. When the data signal has a low signal-to-noise ratio, a derivation mode is adopted to determine the data conversion edge, and a large error exists.

The existing clock recovery methods are basically based on a phase-locked loop clock recovery mode, and do not mention a normal frequency clock recovery mode, and extract the time interval error TIE for analyzing the component components containing jitter.

Because the channel generally exhibits low-pass characteristics, signal degradation due to inter-symbol interference caused by channel insertion loss and reflection or crosstalk from neighboring channels and noise from other sources becomes more pronounced, some points in the edge crossings will be lost in the received signal, thereby introducing errors into the conventional clock recovery based on edge crossing times, and time interval errors extracted from the clock recovered by this means will have a greater impact on jitter decomposition and measurement based on time interval errors.

Disclosure of Invention

The invention aims to solve the problems, and provides a time interval error jitter extraction method based on Hilbert transformation, which is used for carrying out high-precision normal frequency clock recovery on a measured data signal to obtain time interval error information of retaining all jitter components.

The technical scheme adopted by the invention is as follows:

The time interval error jitter extraction method based on Hilbert transform includes:

carrying out level histogram statistics on the measured data signal to determine a decision level;

Performing Hilbert transformation on the detected data signals to obtain pulse widths of the detected data signals;

Performing histogram statistics on the pulse width, and performing rough estimation on the frequency of the recovered clock to obtain a rough frequency value of the recovered clock;

Determining an accurate frequency estimation value of the recovered clock through the coarse frequency value of the recovered clock;

and carrying out normal frequency clock recovery through the accurate frequency estimation value of the recovered clock to obtain normal frequency clock information, and extracting a time interval error by using the normal frequency clock information.

Further, the acquisition of the pulse width comprises the steps of carrying out Hilbert transformation on the detected data signal, deriving an imaginary part of the Hilbert transformation to obtain a plurality of extreme points, wherein the abscissa of the extreme points is time information of jump edges of the detected data signal, and the interval between adjacent jump edges is the pulse width of the current detected data signal.

Further, the pulse width is subjected to histogram statistics to obtain a pulse width histogram, and the recovery clock rate is roughly estimated by using the pulse width histogram to obtain a rough frequency value of the recovery clock.

Further, the coarse estimation includes coarse estimation of the recovered clock rate by a weighted average method.

Further, the coarse frequency value of the recovered clock is taken as a reference, and accurate estimation of the data code element rate is carried out, so that an accurate frequency estimation value of the recovered clock is obtained.

Further, the obtaining of the accurate frequency estimation value of the recovered clock includes:

Calculating the inverse of the coarse frequency value of the recovered clock to obtain the period of the recovered clock, estimating the number of clocks in each pulse width, namely the number of code elements contained in each pulse width, traversing all pulse widths, and obtaining the total number of code elements;

And calculating to obtain the accurate frequency estimation value of the recovered clock according to the difference between the last time jump information and the first time jump information of the tested data signal.

Further, the time interval error acquisition includes:

And carrying out cross point solving on the recovered clock and the decision level, carrying out cross point solving on the tested data signal and the decision level, and obtaining a difference value of the corresponding cross point as a time interval error TIE.

In summary, due to the adoption of the technical scheme, the beneficial effects of the invention are as follows:

when the invention adopts the Hilbert transform-based method to recover the clock, the phenomenon of missing edge crossing points can be effectively solved, and the accurate data code element conversion time can be obtained.

The invention adopts the mode of carrying out Hilbert transformation on data, utilizes the imaginary part of the Hilbert transformation to contain the amplitude and phase information of the data signal, and the extreme point represents the time information of the jump of the data signal, thus being capable of obtaining more accurate data code element width;

The invention provides a mode of coarse frequency estimation and fine frequency estimation of data code element rate, accurately estimates the data code element rate, provides a basis for extracting high-precision time interval error jitter, and improves the accuracy of jitter measurement analysis.

Drawings

FIG. 1 is a functional block diagram of a Hilbert transform-based time interval error jitter extraction method of the present invention;

FIG. 2 is a flow chart of a Hilbert transform-based time interval error jitter extraction method of the present invention;

FIG. 3 is a graph of the data signal under test collected in example 2 of the present invention;

FIG. 4 is an exemplary diagram of the imaginary part of the Hilbert transform of the data signal in embodiment 2 of the present invention;

FIG. 5 is a statistical histogram of pulse width for a conventional method and the method of the present invention;

FIG. 6 is a closed data eye diagram of a constant frequency clock recovery based on conventional histogram pulse width statistics in accordance with example 2 of the present invention;

Fig. 7 is an expanded data eye diagram of the normal frequency clock recovery based on the "coarse frequency estimation+fine frequency estimation" method in embodiment 2 of the present invention.

Detailed Description

The present invention will be described in detail with reference to the accompanying drawings.

The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.

Example 1

The embodiment provides a time interval error jitter extraction method based on Hilbert transform, which comprises the following steps:

Acquiring the data signal comprises the steps of acquiring the measured data signal from acquisition equipment, such as a real-time oscilloscope or an ADC acquisition module, carrying out level histogram statistics on the measured data signal PAM-N, and determining corresponding basic information such as a judgment level.

The imaginary part of Hilbert transformation contains the amplitude and phase information of the measured data signal, the extreme point represents the time information of the jump of the data signal, the time information of all extreme points is extracted, and the time information of the adjacent extreme points is the pulse width of the current data.

And carrying out histogram statistics on all pulse widths, carrying out coarse estimation on the frequency of a recovered clock by adopting a weight average method, namely searching the counted pulse histogram to obtain a first maximum pulse width, taking the derivative of the first maximum pulse width as the coarse frequency estimation of a tested data signal, wherein the method comprises the following specific formula:

(1)

In the formula, To recover a rough estimate of the clock rate,In order to be a pulse width,For statistical probability at the normalized minimum pulse width histogram, i is a natural number,The number of boxes divided at the minimum pulse width histogram;

Further, the coarse frequency value of the recovered clock is inverted to obtain the period of the recovered clock, the number of clocks in each pulse width, namely the number of code elements contained in each pulse width is estimated, all pulse widths are traversed to obtain the total number of code elements, and the accurate frequency estimated value of the recovered clock is calculated according to the difference between the last time jump information and the first time jump information of the tested data signal, wherein the accurate frequency estimated value is calculated according to the following formula:

(2)

In the formula, To recover an accurate frequency estimate of the clock signal,、The time of the jump of the last data signal and the first data signal obtained based on Hilbert transformation;

And extracting the normal frequency clock according to the accurate estimated value of the recovered clock frequency, and respectively obtaining the intersection point of the recovered clock and the decision level and the intersection point of the measured data signal and the decision level by using an interpolation algorithm, wherein the interpolation of the corresponding intersection point is the time interval error.

And carrying out code element cutting on the data code element according to the extracted normal frequency clock information, displaying an eye pattern, and calculating related eye pattern information, wherein the related eye pattern information comprises basic information such as the eye width, the eye height, the signal to noise ratio and the like of the data signal.

Example 2

As shown in fig. 1, in the present embodiment, a functional block diagram of a time interval error jitter extraction method based on hilbert transform is shown, in this embodiment, at a data acquisition module 100, a measured data signal, for example, a measured data signal obtained by waveform drawing shown in fig. 3, is acquired, a clock recovery unit module 105 is used to extract a symbol rate in the measured data signal, that is, a coarse/precise frequency of a recovery clock, by using the high-precision recovery clock, further, the measured data signal may be subjected to symbol cutting, and the measured data signal may be subjected to recovery decoding by a recovery data module 110, or may also be subjected to eye diagram display and measurement by an eye diagram measurement module 115, and a TIE extraction unit module 120 may be used to implement extraction of a data TIE according to the high-precision recovery clock, and perform jitter measurement analysis including all jitter component.

Referring to fig. 2, a flowchart of a specific implementation of the method of this embodiment is shown, after a measured data signal is obtained, the corresponding basic information such as level decision reference is determined through level histogram statistics, the measured data signal is subjected to hilbert transform to obtain time information of a jump edge of the measured data signal, the accurate symbol rate of the measured data signal is extracted by adopting a method of coarse frequency estimation and fine frequency estimation, high-precision normal frequency clock recovery is performed, high-precision TIE extraction is performed according to the recovered clock, and further, symbol cutting can be performed on the measured data signal by utilizing rising and falling edge information of the recovered clock, so as to obtain basic information such as eye width, eye height, signal to noise ratio and the like of the data signal.

As shown in fig. 3, which depicts the characteristics of a signal with high intersymbol interference, the signal does not reach a stable final value in a short time due to the low-pass filtering characteristics of the transmission line, resulting in the amplitude of the high-speed data signal varying with the bit pattern, the amplitude of the data signal eventually rises to a steady-state high level in a long series of "1", as shown at a point a in fig. 3, and reaches its true low level in a long series of "1" s, as shown at a point B in fig. 3, because the previous series of "1" s makes the initial level of the current "0" relatively high, the "0" undergoes a relatively long switching time to reach the threshold value, and a relatively long time to reach its true low level. Therefore, the time that the "0" crosses the threshold lags the ideal time and the amplitude is relatively high, and the next "1" such as point C in FIG. 3 causes the "0" signal at point B to switch back again before reaching its ideal low level, so the "1" signal is very close to its threshold value and the time that it crosses the threshold is advanced from the ideal time. Therefore, the conventional pulse width statistical graph based on the edge crossing method has the problems of edge deletion, too small pulse width statistics and the like, so that serious distortion of data code element rate estimation is caused, the data pulse width statistics is realized by adopting the Hilbert transform method with edge information identification, as shown in fig. 4, the imaginary part of the Hilbert transform can well identify the time information of data signal edge conversion, more accurate pulse width statistics is obtained, and compared with the method for acquiring data signal edge conversion by the conventional spectral line method, the method has better noise tolerance.

In order to obtain the symbol rate of the measured data signal, the conventional method is shown in fig. 5, the first peak position in the histogram is 1 UI wide, the second peak position is 2 UI wide, and so on, so the symbol width of the data signal is t=l1+l2/2+l3/3+l4/4, and there is a large error in the data symbol rate estimated by this method, and when the normal frequency clock recovered according to this rate performs symbol slicing, the phenomenon that the data eye is closed is seriously caused, as shown in fig. 6. Furthermore, if the TIE extracted according to the normal frequency clock of the recovery method also causes larger error and cannot perform correct jitter measurement analysis, the method of the embodiment provides a method of coarse frequency estimation and fine frequency estimation to estimate the code element rate to obtain a high-precision normal frequency recovery clock, and extracts precise TIE data to perform jitter measurement analysis, and when the normal frequency recovery clock obtained by the method of the embodiment cuts a data code element, the obtained eye pattern is obviously opened, as shown in fig. 7, so as to obtain more accurate basic information such as the eye width, eye height, signal to noise ratio and the like of the data signal.

In summary, the invention adopts the Hilbert transform and has the advantages that when the transmitted channel quality is poor, higher ISI and noise possibly exist in the data signal, so that the data is converted to the next symbol level when the level decision reference is not passed, when the clock recovery is carried out by adopting the edge crossing method, the phenomenon of missing edge crossing points can occur, the symbol width obtained by adopting the edge crossing method is obviously smaller than the nominal symbol width of the data, and when the clock recovery is carried out by adopting the Hilbert transform-based method, the phenomenon of missing edge crossing points can be effectively solved, and the accurate data symbol conversion time can be obtained.

When a data signal has high noise, the traditional spectral line method based on the derivative mode has noise sensitivity, high-frequency components and noise in the signal can be amplified, and even slight noise can cause false edge detection or false detection. If jitter is present in the signal, even small changes in frequency or phase can result in loss or blurring of edge information. In the invention, the Hilbert transformation is carried out on the data, the imaginary part of the Hilbert transformation contains the amplitude and phase information of the data signal, the extreme point represents the time information of the jump of the data signal, and the more accurate data code element width can be obtained;

The traditional estimation based on the data code element width adopts a method of searching the maximum value of pulse width histogram statistics, and has larger error when estimating the accurate data code element rate, so that not only can the code element cutting error taking a recovered clock as a reference be caused, and the condition that an eye diagram is closed, but also the jitter decomposition measurement based on the TIE can cause larger error, and the error of the feedback design directivity of a high-speed serial link can be caused; the invention provides a mode of coarse frequency estimation and fine frequency estimation of data code element rate, accurately estimates the data code element rate, provides a basis for extracting high-precision time interval error jitter, and improves the accuracy of jitter measurement analysis.

The principles and embodiments of the present invention have been described herein with reference to specific examples, which are intended to be merely illustrative of the methods of the present invention and their core ideas. It should be noted that it will be apparent to those skilled in the art that various modifications and adaptations of the invention can be made without departing from the principles of the invention and these modifications and adaptations are intended to be within the scope of the invention as defined in the following claims.