VNA不确定度传播的协方差矩阵分析及在功率计校准中应用

Vector network analysis is currently the essential technique for the RF microwave components. According to the measurement and calibration features of these network analyzers, this paper proposed a covariance-matrix-based uncertainty analysis procedure and algorithm. Take full account of the complex frequency information between the imaginary and the real parts, the different frequency, the correlation between different variables, the method can effectively analyzer the quantitative calculation of systematic and random uncertainty, and to ensure the correctness and accuracy of a multi-variant uncertainty analysis. Complying with the functional relationships in practice, this paper used the power-meter calibration experiment, and analyze the calculation of quantitative uncertainty propagations from VNA calibration standards to S-parameter measurements and finally to the power-meter calibration results, and used Monte Carlo simulation to test the correctness of the method.