RL-Based Digital Pre-distortion for Drive Signals in Non-Differentiable Channel
Arash Rabiepoor, Leslie Ann Rusch, Ming Zeng · 2024
This paper proposes a novel digital pre-distortion based on deep reinforcement learning to combat instantaneous nonlinearities in electrical back-to-back communication systems. Simulation results demonstrate its superiority over the case without DPD in terms of bit error rate performance.