Reducing Power Consumption of Digital Predistortion for RF Power Amplifiers Using Real-Time Model Switching

In this article, we propose a new behavioral modeling method to reduce the running complexity and power consumption of digital predistortion (DPD) models for radio frequency amplifiers. By employing proposed method, different cross terms in DPD model can be switched dynamically real time. Each cross-term branch further choose coefficients from multiple coefficient sets, which improves performance with little extra complexity. The switch both branches is realized using decision tree-based controller, leading very low run-time optimizing selection process, input samples most suitable configuration that contributes linearization performance. As only one activated at time, greatly reduced. Based on experimental results, achieve excellent significantly reduced consumption.