| With the rapid development of modern electronic information technology,the application of high-speed signal sampling is more and more extensive.For example,in the broadband test system,radar,telemetry remote sensing,image processing,broadband analog IC and other areas need to use high-speed data acquisition system.However,according to the Nyquist principle,in order to rebuild the signal completely,the sampling frequency is at least twice the signal frequency.In fact,in the field of high-speed signal,single-chip ADC does not achieve such a high sampling rate.So in the 20 th century,80 years,some researchers put forward the time interleaved sampling structure to effectively solve this problem,multi-channel parallel sampling system is the use of multi-chip low-speed ADC to form a time-interleaved structure of the signal sampling,and in the back-end combination,Thus achieving high-speed sampling of the signal.The actual TIADC(time-interleaved analog-to-digital converter)system causes the sampled signal to form a non-uniform sampling waveform due to the mismatch error(offset error,gain error,time error)between the channels Introduced a messy spectrum.Therefore,in order to completely rebuild the signal,we must first estimate the mismatch error between the signals,and then these errors are effectively calibrated.The paper mainly studies the two aspects of channel error estimation and error calibration.Considering that there are some shortcomings in the existing methods,this paper has improved the estimation and correction algorithms in mismatch error,and has made more in-depth Research.(1)The estimation of the mismatch error has been improved.The classical sine fitting method requires most knowledge of the frequency of the signal in advance when the error estimate is made.In most applications,the frequency of the signal can not be accurately known.Therefore,this paper has a "two-step decomposition" of the estimation of mismatch errors.Firstly,the frequency of the signal is estimated accurately by using the fast Fourier transform based on the phase difference,and the error of the signal is estimated by the classical sinusoidal fitting method.Combine these two algorithms so that the exact mismatch error estimates can be made without knowing the signal frequency.(2)The cubic spline interpolation method is used to calibrate the time base error,and the operation is simplified according to the periodic characteristics of the post-sampling signal.In the case of sampling only the time base error,the signal reconstruction of the cubic spline interpolation method is described in detail,and the cubic spline interpolation method is improved by using the periodicity of the periodic nonuniform sampling.the complexity.The theoretical reconstruction of the fractional delay filter is done,and the cubic spline interpolation method and the fractional delay filter method are compared.In order to prove the effectiveness of the improved algorithm,the paper takes matlab simulation.The simulation results show that the combination of the two algorithms is feasible for the estimation of the channel mismatch error and the accuracy of the error estimation is high.In the aspect of signal reconstruction,the improved cubic spline interpolation method has been improved obviously in the non-spurious dynamic range and signal-to-noise ratio. |
