Clipping-and-Bank-Filtering Technique in Joint Crest Factor Reduction and Digital Predistortion for Power Amplifiers

Crest factor reduction (CFR) and digital predistortion (DPD) are often used to enhance power amplifiers (PA) power efficiency while compensating for its nonlinearity. Among the joint CFR/DPDs in the literature, the one based on a parametric model has advantages on running complexity because the CFR and the DPD can be merged into a single model in the implementation. The model coefficients estimation is based on a reference correction signal computed by clipping the predistorted signal. In this paper, we propose a clipping-and-bank-filtering (CABF) method for generating the reference correction signal. The performance obtained by the proposed CABF approach outperforms the conventional clipping methods in terms of adjacent channel power ratio at PA output. Its effectiveness is validated by simulation results on a Wiener model Power Amplifier with a Long Term Evolution-Advanced signal.