Reduced Uncertainty of Measurement for Vector Network Analysers Employing Regression During Calibration

We formulate the calibration process for a one-port vector network analyser (VNA) as a regression problem making full use of the uncertainty information available for the indicated and 'true' voltage reflection coefficients (VRCs) of the standards. The benefit of over-determining the calibration in terms or reduced uncertainties is demonstrated for VRC measurements at RF frequencies