Neural Network Based DPD for Fully-Connected Hybrid Beamforming mMIMO Transmitters

In this paper, a neural-network (NN) based digital predistortion (DPD) model is proposed for the fully-connected (FC) hybrid beamforming (HBF) massive MIMO (mMIMO) transmitters. Comparing to existing models, the proposed model can effectively linearize the virtual main far-field signal and achieves a better performance. Simulations on a 2-stream 64-element FC HBF array demonstrate the effectiveness of the proposed DPD models.