Towards FPGA Implementation of Neural Network-Based Nonlinearity Mitigation Equalizers in Coherent Optical Transmission Systems

Freire, Pedro J.; Anderson, M.; Spinnler, Bernhard; Bex, Thomas; Prilepsky, Jaroslaw E.; Eriksson, Tobias A.; Costa, Nélson; Schairer, Wolfgang; Blott, Michaela; Napoli, Antonio; Turitsyn, Sergei K.

doi:10.48550/arxiv.2206.12180

Cited by 1 publication

(1 citation statement)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Figures 4 and 5 demonstrate that the use of more than 3 channels contributes less to the improvement of the BER. Although it is out of the scope of this study, the hardware implementation of RNN equalizers in optical communication systems is a challenge that already concerns the research community [31]. We believe that the DSP implementation of a three-channel bi-RNN equalizer requiring a rather reasonable number of h.u.…”

Section: Complexity Analysismentioning

confidence: 99%

Performance and Complexity Evaluation of Recurrent Neural Network Models for Fibre Nonlinear Equalization in Digital Coherent Systems

Deligiannidis

Mesaritakis

Bogris

2020

2020 European Conference on Optical Communications (ECOC)

View full text Add to dashboard Cite

Kerr nonlinearity in the form of self-and crossphase modulation imposes a fundamental limitation to the capacity of wavelength division multiplexed (WDM) optical communication systems. Digital back-propagation (DBP), that requires solving the inverse-propagating nonlinear Schrödinger equation (NLSE), is a widely adopted technique for the mitigation of impairments induced by Kerr nonlinearity. However, multichannel DBP is too complex to be implemented commercially in WDM systems. Recurrent neural networks (RNNs) have been recently exploited for nonlinear signal processing in the context of optical communications. In this work, we propose multi-channel equalization through a bidirectional vanilla recurrent neural network (bi-VRNN) in order to improve the performance of the single-channel bi-VRNN algorithm in the transmission of WDM M-QAM signals. We compare the proposed digital algorithm to full-field DBP and to the single channel bi-RNN in order to reveal its merits with respect to both performance and complexity. We finally provide experimental verification through a QPSK metro link, showcasing over 2.5 dB optical signal-to-noise ratio (OSNR) gain and up to 43% complexity reduction with respect to the single-channel RNN and the DBP.

show abstract

Section: Complexity Analysismentioning

confidence: 99%

Performance and Complexity Evaluation of Recurrent Neural Network Models for Fibre Nonlinear Equalization in Digital Coherent Systems

Deligiannidis

Mesaritakis

Bogris

2020

2020 European Conference on Optical Communications (ECOC)

View full text Add to dashboard Cite

show abstract

Towards FPGA Implementation of Neural Network-Based Nonlinearity Mitigation Equalizers in Coherent Optical Transmission Systems

Abstract: For the first time, recurrent and feedforward neural network-based equalizers for nonlinearity compensation are implemented in an FPGA, with a level of complexity comparable to that of a dispersion equalizer. We demonstrate that the NN-based equalizers can outperform a 1-step-per-span DBP.

Cited by 1 publication

References 10 publications

Performance and Complexity Evaluation of Recurrent Neural Network Models for Fibre Nonlinear Equalization in Digital Coherent Systems

Performance and Complexity Evaluation of Recurrent Neural Network Models for Fibre Nonlinear Equalization in Digital Coherent Systems

Contact Info

Product

Resources

About