An Adaptive Method for Linearizing Wideband Power Amplifiers with Band-Limited Feedback Path

Abstract

This paper proposes a Simultaneous Perturbation Stochastic Approximation (SPSA) algorithm based on dualobjective optimization of Normalized Mean Squared Error (NMSE) and Adjacent Channel Power Ratio (ACPR) called SPSA-DO. Since NMSE is mainly affected by in-band distortion and less by out-of-band distortion, the SPSA-DO algorithm can iteratively minimize NMSE and ACPR simultaneously in the time and frequency domains, respectively. Additionally, using a small number of FFT bins close to the desired channel in the ACPR calculation allows the use of a lowpass filter (LPF) in the feedback path to limit the bandwidth, thereby reducing the feedback sampling rate of analog-to-digital converter (ADC) and current consumption. Simulation results show that the SPSA-DO algorithm converges faster than the existing SPSA algorithm in reducing NMSE and ACPR.