Abstract :
Stringent contractual specifications covering complex radar, communications, and space probe systems frequently call for accuracies beyond those derived from existing standards. In many cases, the true value of a standard is not exactly known: in addition, uncertainties are incurred during a measurement process. These, along with the hierarchy of calibrations from higher-to lower-echelon testing laboratories, make it difficult to report separately the inaccuracies of the standard and the measurement system. In some instances, the accuracy of a measurement system and a standard is stated conservatively. This is usually justified pending historical information from data for determining stability characteristics. Statistical tools are available, however, for use in determining inaccuracies and thus improving confidence in accuracy statements. These statistical tools are used in evaluating insertion loss measurements of commercially available fixed-attenuation standards at microwave frequencies where system stability, alignment and waveguide flexure of components are of major concern.
