Behavioural models of injection-locked self-oscillating microstrip antennas for multimedial communication systems

The simulation of integrated self-oscillating microstrip antennas injection-locked by digitally modulated RF signals is a complex task requiring the concurrent use of electromagnetic analysis, envelope-oriented harmonic balance, and bifurcation theory. The paper discusses the development of a class of neural behavioural models that can be used to accurately and efficiently describe these subsystems in a way suitable for large-size system simulation.