Developing Frequency-domain Models for Nonlinear Circuitsbased on Large-signal Measurements*

We describe an approach for generating artificial neural network (ANN) models for nonlinear devices, based upon frequency-domain data measured using a nonlinear vector network analyzer (NVNA). We demonstrate that these ANN models give accurate descriptions of the input-output relationships of the device over the span of the measurements. We also use the models to construct application-specific engineering figures of merit using appropriate ratios of wave variables measured under large-signal conditions. We obtain an independent check by comparing diode-circuit models, generated by means of this methodology, to a compact model simulated in commercial harmonic-balance software.