Two-Input Nonlinear Dynamic Model Inversion for the Linearization of Envelope-Tracking RF PAs

We present an algorithm for the real-time inversion of a two-input behavioral model applicable to supply-modulated radio-frequency (RF) power amplifiers (PAs). This approach includes the nonlinear dynamic effects on the RF output due to the mutual interactions between the modulated supply and the RF input. The model inversion consists of an iterative procedure that only relies on the model, which has been empirically identified. It results in an open loop linearization algorithm fully implementable in a digital architecture. The method allows exploiting the trade-off between the power-added efficiency (PAE) and the digital predistortion (DPD) computational load for a given linearity target.