Matlab scripts which perform DPD with real-valued feedback (only real part or imaginary part of IQ demodulator output). Firstly, the forward PA model is estimated and subsequently it is used to train DPD coefficients using the indirect learning architecture.
The scripts supplement a research paper Digital Predistorter with Real-Valued Feedback Employing Forward Model Estimation by Jan KRAL, Tomas GOTTHANS, Roman MARSALEK, and Michal HARVANEK.
RUN_ANALYSIS_01.m - main script, runs all the analysis presented in the paper. During the analysis, the main charts are plot.
Run_Diff_DPD_Arch.m - script which executes different DPD architectures including the proposed technique using the forward model extraction and ILA
PLOT_01.m - plots results in the form presented in the paper
If used for research, plese cite our paper