Density estimation models for strong nonlinearities in RF power amplifiers

The paper presents a method for modeling strong nonlinear effects in power amplifiers based on the principles of density estimation. The static nonlinear transfer function is obtained by averaging measured data. The performance obtained with density estimation is similar to the one using high order nonlinear static polynomial models. The benefit of considered method over the ones using polynomial models is that the former estimates blindly the structure of the transfer function and does not suffer from numerical instabilities.