Efficient Training of Volterra Series-Based Pre-distortion Filter Using Neural Networks

TL;DR

This paper introduces a neural network-based direct learning method for training Volterra series pre-distortion filters, demonstrating superior performance in simulating high-speed 64-QAM transmitters under nonlinear and noisy conditions.

Contribution

The paper proposes a novel direct learning approach for Volterra series pre-distortion filters using neural networks, improving training efficiency and performance.

Findings

Outperforms conventional training methods in simulations

Effective under varying nonlinearity and noise conditions

Applicable to high-speed digital transmitters

Abstract

We present a simple, efficient "direct learning" approach to train Volterra series-based digital pre-distortion filters using neural networks. We show its superior performance over conventional training methods using a 64-QAM 64-GBaud simulated transmitter with varying transmitter nonlinearity and noisy conditions.