Novel approach for static nonlinear behavior identification in RF power amplifiers exhibiting memory effects

In this paper, an experimental approach is proposed to accurately identify, under a modulated signal drive, the memoryless nonlinearity of power amplifiers exhibiting memory effects. It is experimentally demonstrated that, when they are present, memory effects bias the extracted static nonlinearity. Accordingly, the sampling rate of the WCDMA test signal waveform is varied to reduce the signal's bandwidth. It is shown that this approach minimizes the memory effects contribution to the amplifier's nonlinear behavior and leads to accurate characterization of the 'true' static nonlinearity. The performance of the proposed approach is then assessed through experimental memoryless digital predistortion.