Noise Parameters Extraction with Special Source Reflection Coefficient Selection Patterns

Source reflection coefficient selection patterns have important influence on noise parameters extraction based on least-squares method. As an alternative method, genetic algorithm is able to extract noise parameters without linearization. This paper discusses if source reflection coefficient selection has no or less influence on noise parameters extracted by genetic algorithm. Two special source reflection coefficient selection patterns are used to extract noise parameters of a low noise microwave amplifier. Experiment results show that source reflection coefficient selection does have influence on noise parameters extracted by genetic algorithm and inappropriate selection patterns could degrade extraction results.