Wiener PA model can be perfectly predistorted by Hammerstein model. According to the structural characteristics of these two models

A separation predistortion method was proposed. The method first separated memory subsystem and nonlinear subsystem from Wiener PA

and then got their corresponding predistortion subsystems by identification algorithm. At last

it combined the two subsystems to a Hammerstein system and then linearized Wiener PA. The method avoied interference between memory and nonlinear factors

This kind of interference can be found in the former methods which is not only complexity

but also leads to slow convergence speed and limited convergence precision. Simulation is made to signify that the method proposed is simple and can get more fast speed and more accurate convergence result.