Band-limited Adaptation for DPD Systems with Low-pass Filtering at the Feedback Path

This paper explores an adaptation algorithm for DPD systems with intuitive direct learning that takes into account digital low-pass filtering at the feedback path. In scenarios where the low-pass filter is narrow, the conventional DPD experiences performance degradation as it aims to linearize in-band distortion while minimizing the out-of-band power of the pre-distorted signal, potentially improving DPD stability. In contrast, the proposed band-limited adaptation approach focuses solely on in-band linearization, achieving comparable performance to DPD systems with wideband feedback filters. Moreover, the band-limited adaptation method outperforms the conventional DPD, demonstrating a gain of 2 ~ 3 dB as observed in the test cases.