Digital predistortion based on Zernike polynomial functions for RF nonlinear power amplifiers

Predistortion linearizer for high nonlinear power amplifiers (PAs) is an important topic of research in wireless applications. Behavioral models based on polynomial functions are widely used for both PAs and predistorters. However, they often suffer from numerical instability problems when we include high nonlinear-order in the mathematical model. In this work, we propose a new predistorter model considering the use of an orthogonal basis. This basis is composed of Zernike polynomials, which are very robust and stable when they are applied to classical identification techniques. Computer simulations in a typical DVB-T OFDM transmission confirm the predistorter computed with the Zernike polynomial model provides improvements in adjacent channel power ratio with stability and moderate complexity in the DPD coefficient estimation.