Revisiting Efficient Multi-Step Nonlinearity Compensation With Machine Learning: An Experimental Demonstration

Abstract

Efficient nonlinearity compensation in fiber-optic communication systems is considered a key element to go beyond the "capacity crunch". One guiding principle for previous work on the design of practical nonlinearity compensation schemes is that fewer steps lead to better systems. In this paper, we challenge this assumption and show how to carefully design multi-step approaches that provide better performance-complexity tradeoffs than their few-step counterparts. We consider the recently proposed learned digital backpropagation (LDBP) approach, where the linear steps in the split-step method are re-interpreted as general linear functions, similar to the weight matrices in a deep neural network. Our main contribution lies in an experimental demonstration of this approach for a 25 Gbaud singlechannel optical transmission system. It is shown how LDBP can be integrated into a coherent receiver DSP chain and successfully trained in the presence of various hardware impairments. Our results show that LDBP with limited complexity can achieve better performance than standard DBP by using very short, but jointly optimized, finite-impulse response filters in each step. This paper also provides an overview of recently proposed extensions of LDBP and we comment on potentially interesting avenues for future work.

Advanced Convolutional Neural Networks for Nonlinearity Mitigation in Long-Haul WDM Transmission Systems

“…We assume here that signals at each channel propagate at the carrier frequency and after the linear step group delay corresponding to channel frequency and distance propagated is compensated. To account for the group delay difference, a realvalued fractional delay (FD) filter for each spectral channel can be used after convolutional layer at each linear step [25], [26]. Furthermore, to reduce the computational complexity, we set the step length so that the time delay for each channel is divisible by the duration of the symbol interval T , as it was suggested in [20].…”

Evaluation of Frequency-Domain Learned Digital Back-Propagation Nonlinear Compensation for Unrepeatered Optical Links

“…Although [4] has demonstrated that joint optimization of the filters can make time-domain filtering effective for LDBP, with a desirable complexity reduction for hardware implementations, we have chosen to restrict the analysis of this work to frequency-domain implementation of LDBP, so its performance can be directly compared with the equivalent untrained conventional DBP, with no discussion on the complexity of the different architectures. This choice of implementation also allows us to use unsupervised learning, which greatly simplifies the training procedure and improves robustness of possible real-time implementations.…”

“…Noticing the structural similarity between DBP and artificial neural networks (ANN), recent works [3,4] have proposed combining data-driven optimization techniques developed for ANN and the structure of DBP, creating an architecture that Lucas Silva Schanner, Optical Communication Solutions, CPQD, Campinas-SP, e-mail: schanner@cpqd.com.br;…”

A Layer-Reduced Neural Network Based Digital Backpropagation Algorithm for Fiber Nonlinearity Mitigation

“…A variant of the NN-based NLC algorithm, called learned DBP (LDBP), treated the linear steps and the nonlinear steps of SSFM as the linear layers and the nonlinear activation functions of an NN [15], [16]. LDBP showed a significant performance gain compared with the DBP in the experimental settings [17]- [19]. The taps in linear layers can be extremely short by gradually pruning them during training [17].…”

“…LDBP showed a significant performance gain compared with the DBP in the experimental settings [17]- [19]. The taps in linear layers can be extremely short by gradually pruning them during training [17]. In this case, the convolutions of linear layers can be performed in the time domain for low power consumption [20].…”