Robust and fast iterative algorithm based on Levenberg-Marquardt and spectral extrapolation for wideband digital predistortion of RF power amplifiers

Abstract

The analog-to-digital converter (ADC) is one of the most expensive components in an RF transmitter, conventional digital predistortion (DPD) could be seriously limited in dealing with the wideband feedback signals. Modern wireless communication systems are requiring higher data rate and wider bandwidth, a more powerful linearization technique is required. In this paper, we propose a robust iterative DPD coefficients extraction algorithm based on Levenberg-Marquardt algorithm and spectral extrapolation. By applying this method, the speed of ADC can be reduced significantly and thus save the cost. Experimental results demonstrate that the proposed method is numerically stable and more suitable for iteration than the damped Gauss-Newton method mentioned in [7].