Application of neural networks in microwave FET transistor noise modeling

An application of neural networks in microwave FET transistor (MESFET) noise modeling is presented. A multilayer perceptron neural network is implemented to model four transistor noise parameters. Inputs for the neural models include small-signal intrinsic equivalent circuit elements, two equivalent temperatures and frequency, while outputs are noise parameters. Pospieszalski's approach is employed to characterize the noise parameters.