Reduced-Complexity Digital Predistortion: Adaptive Spline-Based Hammerstein Approach

In this paper, a novel digital predistorter concept for power amplifier (PA) linearization is proposed, with particular emphasis on reduced processing complexity in future 5G and beyond wideband radio systems. The proposed method builds on a complex spline interpolated look-up table (LUT) followed by a linear finite impulse response (FIR) filter, comprising essentially a Hammerstein-type nonlinear system. For reliable parameter learning, gradient-descent based adaptive learning rules are derived, allowing for the estimation of the spline control points and the FIR filter parameters in a decoupled manner. Large set of experimental results are provided, with specific focus on 5G New Radio (NR) systems, showing successful linearization of multiple sub-6 GHz PA samples as well as a 28 GHz active antenna array, incorporating channel bandwidths up to 200 MHz. Explicit performance-complexity comparisons are also reported against two known reference solutions, namely a memory polynomial (MP) based DPD and a linear interpolated LUT. The results show that the linearization performance of the proposed method is very close to that of a memory polynomial while clearly outperforming the linear interpolated LUT. Additionally, it is shown that the processing complexity of the proposed DPD is commonly some 60 % lower than that of the MP based DPD, offering thus a very good complexity-performance tradeoff.