A study of uncertainty estimation for time-domain analysis by considering incompleteness of TRL calibration kit

Abstract

We have developed a new algorithm of the Monte-Carlo calculation for the uncertainty evaluation of time-domain measurements, which directly relates to the incompleteness of a calibration standard used on a vector network analyzer. In contrast to the previous study, the propagation of S-parameter uncertainty is considered, and correlations between measured S-parameters at different frequency points are rigorously taken into account, which is important because the time-domain analysis treats the frequency dependence of the S-parameter. We confirmed that the uncertainty of the peak determination at the time-domain is strongly affected by S-parameter uncertainties especially when we measure time-domain signals including multiple peaks, and quantitatively clarified that it is essential for the uncertainty analysis of time-domain measurements to evaluate the propagation of S-parameter uncertainties appropriately.