Abd Elmonieum Abd Elbary
Neural Based Prediction of Scattering and Noise Parameters for Solid State Microwave Transistors
Abstract – In this paper a Neural based algorithm is designed to predict the scattering and noise parameters for solid state microwave devices with application to the High Electron Mobility Transistor ( HEMT). This makes use of a finite number of measurements to extend the scattering and noise parameters to a wide range of applied voltages, currents and operating frequencies. The algorithm used is based on the feed-forward technique. Results show good agreement with the measured parameters which is adequate for the design of small and large signal amplifiers. Keywords: S-parameters, Noise parameters, Neural network, HEMT