Orthonormal-basis power amplifier model reduction

A new power amplifier behavioral model is proposed in this paper. The memory polynomial model is combined with a Kautz-Volterra like filter to construct this novel parallel three-box model. With this technique, the number of parameters required to obtain a certain accuracy can be reduced, and a higher nonlinear order and longer memory depth can be modeled compared to the Kautz-Volterra model. Results from modeling a power amplifier are presented and compared to other models.