Bi-directional gated recurrent unit neural network based nonlinear equalizer for coherent optical communication system

Liu, Xinyu; Wang, Yongjun; Wang, Xishuo; Xu, Hui; Li, Chao; Xin, Xiangjun

doi:10.1364/oe.416672

Cited by 91 publications

(33 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Thus, the NN equalizer, which is reverting the channel nonlinear effects, continues to function well for other modulation formats. This is in stark contrast to the case of classification equalizers (classifiers) [46][47][48], because the latter incorporates the decision boundaries in the NN structure itself. For the classifiers, the S-NN will not work with the new target task since its output stage does not capture the different symbol alphabet.…”

Section: Transfer Learning For Different Modulation Formatsmentioning

confidence: 99%

Transfer Learning for Neural Networks-Based Equalizers in Coherent Optical Systems

Freire

Abode

Prilepsky

et al. 2021

J. Lightwave Technol.

View full text Add to dashboard Cite

In this work, we address the question of the adaptability of artificial neural networks (NNs) used for impairments mitigation in optical transmission systems. We demonstrate that by using well-developed techniques based on the concept of transfer learning, we can efficaciously retrain NN-based equalizers to adapt to the changes in the transmission system, using just a fraction (down to 1%) of the initial training data or epochs. We evaluate the capability of transfer learning to adapt the NN to changes in the launch power, modulation format, symbol rate, or even fiber plants (different fiber types and lengths). The numerical examples utilize the recently introduced NN equalizer consisting of a convolutional layer coupled with bi-directional long-short term memory (biLSTM) recurrent NN element. Our analysis focuses on long-haul coherent optical transmission systems for two types of fibers: the standard single-mode fiber (SSMF) and the TrueWave Classic (TWC) fiber. We underline the specific peculiarities that occur when transferring the learning in coherent optical communication systems and draw the limits for the transfer learning efficiency. Our results demonstrate the effectiveness of transfer learning for the fast adaptation of NN architectures to different transmission regimes and scenarios, paving the way for engineering flexible and universal solutions for nonlinearity mitigation.

show abstract

Section: Transfer Learning For Different Modulation Formatsmentioning

confidence: 99%

Transfer Learning for Neural Networks-Based Equalizers in Coherent Optical Systems

Freire

Abode

Prilepsky

et al. 2021

J. Lightwave Technol.

View full text Add to dashboard Cite

show abstract

“…Thus, the NN equalizer, which is basically reverting the channel nonlinear effects, continues to function as well for other modulation formats. This is in stark contrast to the case of classification equalizers (classifiers) [43][44][45] because classification equalizers incorporate the hard-decision boundaries into the NN structure itself. For the classifiers, the S-NN will not work with the new target task since its output stage will not capture the different symbol alphabet.…”

Section: Transfer Learning To Different Modulation Formats Scenariosmentioning

confidence: 97%

“…Furthermore, the re-training process needs much less data, with a reduction of up to 99% of the required training data for the 3 considered symbol rates. The number of required epochs for the TWC fiber link(Case II), decreased by 92%, 73%, and 75%, respectively, when switching to (d) 45 GBd, (e) 65 GBd, and (f) 85 GBd. We can also see that the retraining phase needs less information: 95%, 95%, and 90%, respectively.…”

Section: Transfer Learning To Different Symbol Rate Scenariosmentioning

confidence: 99%

Transfer Learning for Neural Networks-based Equalizers in Coherent Optical Systems

Freire,

Abode,

Prilepsky

et al. 2021

Preprint

View full text Add to dashboard Cite

In this work, we address the paramount question of generalizability and adaptability of artificial neural networks (NNs) used for impairment mitigation in optical transmission systems. We demonstrate that by using well-developed techniques based on the concept of transfer learning, we can efficaciously retrain NN-based equalizers to adapt to changes in the transmission system using just a fraction of the initial training data and resources. We evaluate the potential of transfer learning to adapt the NN to changes in the launch powers, modulation formats, symbol rates, or even fiber plant (different fiber types and lengths). In our numerical examples, we consider the recently introduced combined NN equalizer consisting of a convolutional layer coupled with bi-directional long-short term memory (biLSTM) recurrent NN elements. We focus our analysis on long-haul coherent optical transmission systems employing two types of transmission fibers: the standard single-mode fiber (SSMF) and the TrueWave Classic (TWC) fiber. We also underline the specific peculiarities that occur when transferring the learning in coherent optical communication systems. Our results demonstrate the effectiveness of transfer learning for the fast adaptation of NN architectures to different transmission regimes and scenarios, paving the way for engineering flexible universal solutions for nonlinearity mitigation.

show abstract

“…In the whole matching layer, a bi-directional gated recurrent unit network [27] is used to realize parameter transfer, so forward and backward transfer is distinguished during comparison. Assuming that the return result is x β j after comparing the sentence unit σ β j in question Q with the last sentence unit in answer A, then:…”

Section: Multi-strategy Interaction Layermentioning

confidence: 99%

A Dynamic Attention and Multi-Strategy-Matching Neural Network Based on Bert for Chinese Rice-Related Answer Selection

Wang

et al. 2022

Agriculture

View full text Add to dashboard Cite

To allow the intelligent detection of correct answers in the rice-related question-and-answer (Q&A) communities of the “China Agricultural Technology Extension Information Platform,” we propose an answer selection model with dynamic attention and multi-strategy matching (DAMM). According to the characteristics of the rice-related dataset, the twelve-layer Chinese Bert pre-training model was employed to vectorize the text data and was compared with Word2vec, GloVe, and TF-IDF (Term Frequency–Inverse Document Frequency) methods. It was concluded that Bert could effectively solve the agricultural text’s high dimensionality and sparsity problems. As well as the problem of polysemy having different meanings in different contexts, dynamic attention with two different filtering strategies was used in the attention layer to effectively remove the sentence’s noise. The sentence representation of question-and-answer sentences was obtained. Secondly, two matching strategies (Full matching and Attentive matching) were introduced in the matching layer to complete the interaction between sentence vectors. Thirdly, a bi-directional gated recurrent unit (BiGRU) network spliced the sentence vectors obtained from the matching layer. Finally, a classifier was employed to calculate the similarity of splicing vectors, and the semantic correlation between question-and-answer sentences was acquired. The experimental results showed that DAMM had the best performance in the rice-related answer selection dataset compared with the other six answer selection models, of which MAP (Mean Average Precision) and MRR (Mean Reciprocal Rank) of DAMM gained 85.7% and 88.9%, respectively. Compared with the other six kinds of answer selection models, we present a new state-of-the-art method with the rice-related answer selection dataset.

show abstract

Bi-directional gated recurrent unit neural network based nonlinear equalizer for coherent optical communication system

Cited by 91 publications

References 30 publications

Transfer Learning for Neural Networks-Based Equalizers in Coherent Optical Systems

Transfer Learning for Neural Networks-Based Equalizers in Coherent Optical Systems

Transfer Learning for Neural Networks-based Equalizers in Coherent Optical Systems

A Dynamic Attention and Multi-Strategy-Matching Neural Network Based on Bert for Chinese Rice-Related Answer Selection

Contact Info

Product

Resources

About