Improved ANN approach to bias-dependent scalable noise modelling of microwave FETs

In the earlier work authors have shown that prior knowledge neural approach can be successfully applied for bias dependent noise modelling of microwave transistors, which is very applicable in the cases where there are no enough data to develop pure black-box neural models. In this paper the mentioned model is extended to a class of devices made in the same technology but differing in the gate width. In that way device scalability is included in the model. The modelling approach is applied to a class of pHEMT devices.