In the evaluation of uncertainty related to complex-valued measurements, coverage factors must be derived in order to account for the probability for the measurand to lie within a given uncertainty region, generally of elliptical form [1], [2]. These coverage factors are dependent on the type of two-dimensional Probability Density Function (PDF) assumed. Traditional approaches assume Bivariate Normal distribution in the real and imaginary components. In this paper a new definition for the PDF is explored, which assumes gaussian distribution for the magnitude and uniform distribution for the phase. This so-called Gaussian Magnitude PDF is compared with the Bivariate Normal distribution in real and imaginary. For both models, coverage factors for the usual confidence level of 95percent are derived.
展开▼
