Performance versus Complexity Study of Neural Network Equalizers in Coherent Optical Systems

TL;DR

This study compares the performance and computational complexity of various neural network architectures for nonlinear channel equalization in coherent optical systems, highlighting the CNN+biLSTM as the most effective in high-complexity scenarios.

Contribution

It introduces a comprehensive complexity-performance comparison of neural network equalizers, including novel combinations like CNN+biLSTM, with a detailed complexity derivation applicable to diverse systems.

Findings

CNN+biLSTM achieves the highest Q-factor improvement of 2.9 dB.

Three-layer perceptron performs best under low complexity constraints.

Complexity analysis is generalizable to various physical and engineering systems.

Abstract

We present the results of the comparative analysis of the performance versus complexity for several types of artificial neural networks (NNs) used for nonlinear channel equalization in coherent optical communication systems. The comparison has been carried out using an experimental set-up with transmission dominated by the Kerr nonlinearity and component imperfections. For the first time, we investigate the application to the channel equalization of the convolution layer (CNN) in combination with a bidirectional long short-term memory (biLSTM) layer and the design combining CNN with a multi-layer perceptron. Their performance is compared with the one delivered by the previously proposed NN equalizer models: one biLSTM layer, three-dense-layer perceptron, and the echo state network. Importantly, all architectures have been initially optimized by a Bayesian optimizer. We present the derivation of the computational complexity associated with each NN type -- in terms of real multiplications per symbol so that these results can be applied to a large number of communication systems. We demonstrated that in the specific considered experimental system the convolutional layer coupled with the biLSTM (CNN+biLSTM) provides the highest Q-factor improvement compared to the reference linear chromatic dispersion compensation (2.9 dB improvement). We examine the trade-off between the computational complexity and performance of all equalizers and demonstrate that the CNN+biLSTM is the best option when the computational complexity is not constrained, while when we restrict the complexity to lower levels, the three-layer perceptron provides the best performance. Our complexity analysis for different NNs is generic and can be applied in a wide range of physical and engineering systems.