A Sample Selection Method for Digital Predistortion with Multiple States

In this paper, a low computational complexity sample selection method (SSM) for digital predistortion (DPD) of power amplifier (PA) with multiple states is proposed. The proposed method only needs to determine whether the samples in other states are duplicated with the selected samples in the first state, thus avoiding the need to repeatedly calculate the Euclidean distance between any two samples in the sample space of each state. The experimental results show that compared with previous SSMs, the proposed method reduces computational complexity by about $\mathbf{6 6 \%}$ and has better performance in the scenario with multiple states.