Affine Projection Algorithm Based Direct Adaptations for Adaptive Nonlinear Predistorters

It has been proven that adaptive nonlinear predistorters based on direct adaptation have many advantages over ones based on indirect adaptation. In this work, we develop two affine projection based direct adaptation algorithms for adaptive nonlinear predistorter: nonlinear filtered-x affine projection algorithm and nonlinear adjoint affine projection algorithm. Compared with available least mean square (LMS) based direct adaptation predistorters, predistorters based the new developed algorithms have the advantages fast convergence and efficient computational complexity. The nonlinear adjoint affine projection algorithm is computationally more efficient than the nonlinear filtered-x affine projection algorithm. Simulation results confirm the better performance of our developed algorithms.