Computation of noise parameters using genetic algorithms

Abstract

In this study, an improved computation method for the noise parameters of a linear noisy two-port network is introduced. A new error function, which considers noise figure and source admittance error simultaneously, is proposed to estimate the four noise parameters. The global optimization of the error function is searched directly by using a genetic algorithm.