Real-Valued Feedback Digital Predistortion Based on QR Decomposition Sample Selection

The digital predistortion technique is effective in balancing the efficiency of the power amplifier and the linearity of the output signal, but it also brings a notable hardware cost and computational effort. In this paper, we propose a new DPD architecture that combines a sample selection technique based on QR decomposition with a digital predistorter over a single feedback signal. The method achieves the goal of reducing the computational effort of DPD parameters while optimizing the DPD system architecture by selecting a representative set of samples on a single feedback signal link. The experimental results show that the DPD structure still has good pre-distortion effect while retaining 38% of the original signal data set.