A parallel-filterbank-based Hammerstein predistorter

A parallel filterbank based Hammerstein predistorter is proposed for improving the compensation performance of the traditional Hammerstein predistorter. A two-stage scheme is used to identify the Hammerstein predistorter, which firstly identify the nonlinear part offline, and then adaptively identify the linear part. Simulation results demonstrate that the performance of the proposed PD is superior to the traditional Hammerstein predistorter, especially for the power amplifier with strong memory effects.