ANN model for AlGaN/GaN HEMTs constructed from near-optimal-load large-signal measurements

AlGaN/GaN HEMTs have a high potential for high-power applications at microwave frequencies. We developed a behavioural model corresponding to the operation condition as how this device will be used in power amplifiers, i.e., at a high output load. The model is based on time-domain large-signal measurements, and the representation format is an artificial neural network (ANN). AlGaN/GaN devices on sapphire are known to be temperature sensitive. Therefore, we also consider the incorporation of the self-heating effect in the behavioural model description.