Modified Least Squares Extraction for Volterra-Series Digital Predistorter in the Presence of Feedback Measurement Errors

Measurement errors (in-phase/quadrature imbalance, dc offset, and nonlinearity) in the feedback path can adversely affect the linearization performance of digital predistorter (DPD) for RF power amplifiers (PAs). In this paper, a generalized analysis for the Volterra-series DPD system is presented in the presence of feedback measurement errors. It shows that the DPD coefficients are biased due to these errors. A modified least squares (MLS) method is then proposed for DPD coefficients extraction, which can eliminate the detrimental effect of feedback measurement errors without using a post-compensator. The proposed MLS method has the advantage of being free of behavioral modeling for the feedback path or the post-compensator. However, it can still achieve comparable performance as the state-of-the-art. The performance of the MLS method is validated with both simulations and experiments. The measurement results show that, when a nonideal feedback path is employed to capture the PA output, the proposed MLS method can still ensure a high linearization performance of the DPD, and the results are nearly the same as that when an ideal feedback path is used.