Implementing Neural Network-Based Equalizers in a Coherent Optical Transmission System Using Field-Programmable Gate Arrays

Abstract

In this work, we demonstrate the offline FPGA realization of both recurrent and feedforward neural network (NN)based equalizers for nonlinearity compensation in coherent optical transmission systems. First, we present a realization pipeline showing the conversion of the models from Python libraries to the FPGA chip synthesis and implementation. Then, we review the main alternatives for the hardware implementation of nonlinear activation functions. The main results are divided into three parts: a performance comparison, an analysis of how activation functions are implemented, and a report on the complexity of the hardware. The performance in Q-factor is presented for the cases of bidirectional long-short-term memory coupled with convolutional NN (biLSTM + CNN) equalizer, CNN equalizer, and standard 1-StpS digital back-propagation (DBP) for the simulation and experiment propagation of a single channel dual-polarization (SC-DP) 16QAM at 34 GBd along 17 × 70 km of LEAF. The biLSTM+CNN equalizer provides a similar result to DBP and a 1.7 dB Q-factor gain compared with the chromatic dispersion compensation baseline in the experimental dataset. After that, we assess the Q-factor and the impact of hardware utilization when approximating the activation functions of NN using Taylor series, piecewise linear, and look-up table (LUT) approximations. We also show how to mitigate the Manuscript

FPGA-based Trainable Autoencoder for Communication Systems

“…In [18], the FPGA implementations of recurrent neural network (RNN) and CNN-based equalizers were presented. As a result, the combination of RNN and CNN provided similar performance as the digital back-propagation (DBP) equalizer while achieving a gain compared with the chromatic dispersion compensation baseline.…”

“…In contrast to most previous works [6]- [8], we do not only propose an optimized implementation of the ANN's forward pass (FP) but also tackle the challenges of implementing the backpropagation algorithm on the FPGA, which enables online retraining on the edge device itself, to adapt for varying channel conditions. A related approach is also presented in [9], but contrary to our work their model is based on the split-step solution of the Manakov-PMD equation instead of an ANN, thus it is not channel-agnostic.…”