Low-Complexity Digital Predistortion Algorithm for Nonlinear Correction of MIMO Systems Under Crosstalk

This letter presents a low-complexity digital predistortion algorithm for nonlinear correction in multiple-input and multiple-output (MIMO) transmitters with nonlinear crosstalk. It utilizes a novel augmented parallel nested quasi-Hammerstein (APNQH) model for system modeling and a single-loop three-step iterative Least Squares (LS)-based algorithm for model coefficient extraction. The model comprises multiple parallel quasi-Hammerstein branches and multiple third-order nonlinear filter branches, aiming to balance complexity and accuracy. Validation is performed by 8 × 8 MIMO system simulations and 2 × 2 transmitter measurements. The results exhibit that the proposed algorithm achieves better linearization performance with reduced complexity compared to benchmark methods.