An Experimental Technique for Calibration Uncertainty Analysis

An experimentally based technique for characterizing calibration uncertainty is presented. The approach described calculates uncertainty metrics at the output of the calibration processing chain as opposed to the input. In doing so, this method replaces the complexities of error propagation with the computational effort associated with performing numerous calibrations. Practical applications are demonstrated for a variety of scenarios focused on one-port calibration, illustrating the versatility of the technique. An open-source implementation has been made publicly available as part of the Python module scikit-rf.