Analysis And Processing Of Digital Signals Eye Diagram With Equivalent-time Sampling

For high-speed serial signals,the real-time oscilloscope is limited by the high-speed ADC sampling rate and have difficulty measuring signals with the bandwidth above 4GHz,while the sampling oscilloscope can realize a broadband signal sampling based on the equivalent-time sampling method.However,the equivalent-time sampling method of the sampling oscilloscope will result in non-periodic sampling of the sampled data.In order to evaluate the reconstructed signal quality through the eye diagram method,it is necessary to study the eye diagram analysis method of the equivalent-time sampling signal.Meanwhile,in order to address the impact of hardware circuit jitter and parameter instability on the reconstructed signal of the sampling oscilloscope,it is necessary to conduct research on the data calibration algorithms of the sampling oscilloscope.The main research contents of this paper are as follows:1.In order to reduce the deviation between the reconstructed signal and the measured frequency domain,the DC signal calibration method based on the least squares method was proposed.The principle of least squares fitting was used to calibrate the amplitude of the reconstructed waveform.The Fourier transform principle was used to calculate the frequency of the reconstructed signal,and a stepwise approach was used to calibrate the trigger pulse delay.After multiple experiments,the DC accuracy error was within 0.2%,and the frequency test error was less than 2%,both errors were within an allowable range.2.In order to realize the characteristic analysis of the eye diagram,in this paper,the eye diagram is formed by the equivalent-time sampling method.By determining the transmission rate of the reconstructed digital signal,a single period of the eye diagram is obtained,and the key parameters such as eye height,eye width and signal-to-noise ratio are measured.Meanwhile,to improve measurement accuracy,the K-Medoids clustering algorithm is used for cluster analysis of the eye diagram’s crossing points and the 0/1 level data sets.To facilitate users in evaluating the performance of the eye diagram,templates of eye diagrams for different signals are established.3.To reduce the effects of noise and jitter in the reconstructed signal,this paper studies filtering algorithms for the reconstructed signal in both time and frequency domains.Firstly,an improved algorithm based on adaptive Gaussian mean filtering was proposed by modifying the mean filtering algorithm.This method reduces the influence of outliers,and can continuously judge and compensate data within the window cache area.Secondly,denoising research was conducted on the reconstructed signal using wavelet denoising algorithm.The results show that both methods have high signal-to-noise ratio and can reflect the details of the data,as well as having good smoothness of the signal.Finally,based on the.Net development platform,software for a broadband sampling oscilloscope was developed.The control,communication,signal acquisition,and display of time domain waveforms,as well as measurement and analysis of eye diagram parameters were implemented.Bi-directional communication between the host computer and the device was achieved via USB interface technology.A hardware platform was set up to test the software developed by our institute,which met the design requirements and operated normally.