Digital Predistortion Parameter Identification for RF Power Amplifiers Using Real-Valued Output Data

Abstract

This brief presents a novel digital predistortion parameter identification technique that requires only the acquisition of either the in-phase (I) or the quadrature (Q) component of the power amplifier output signal. To this end, an approach that allows us to estimate the parameters of a model using only one of the IQ components of the model output is presented. Based on experimental results, it is shown that the proposed real-valued measurements-based technique can offer linearization capabilities similar to its complex-valued counterparts. The experimental results also indicate that the proposed technique can be used in combination with other techniques that focus on reducing the speed of analog-to-digital converters.