RF Behavioural Modelling using Multisine Vector Data

D. Schreurs (Belgium)


Nonlinear systems Modelling Identification Microwave devices Artificial neural networks Memory effects


Behavioural models for RF devices are typically based on single-tone data. This type of excitation is however not very representative, as RF devices, being used in telecommunication systems, are usually subjected to modulated excitations. In this work, we show that the use of modulated excitations renders the modelling more efficient from both the experiment design and data handling points of view, and this without loss of accuracy. Also, as modulated excitations can reveal slow-memory effects present in the device, we propose a way to incorporate those in the behavioural modelling description.

Important Links:

Go Back