Evaluating Cross- feature Trained Machine Learning Models for Estimating QoT of Unestablished Lightpaths

Usmani, Fehmida; Khan, Ihtesham; Siddiqui, Mehek; Khan, Mahnoor; Bilal, Muhamamd; Masood, Muhammad Umar; Ahmad, Arsalan; Shahzad, Muhammad; Curri, Vittorio

doi:10.1109/icecce52056.2021.9514154

Cited by 3 publications

(3 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several ML-based QoT estimators have been proposed in the literature [11][12][13][14][15][16][17][18][19]. Good overviews of the most recent ML models and tools developed for lightpath QoT estimation can be found in [20][21][22].…”

Section: B Qot Estimation: Do We Need Neural Network?mentioning

confidence: 99%

“…Several recent studies show a tendency towards favorable results for ANN based estimators [13][14][15][16][17][18][19].…”

Section: Neural Network For Lightpath Qot Estimationmentioning

confidence: 99%

“…In [18], it was possible for an ANN-based model to predict the SNR with a standard deviation of the SNR estimation error of 0.13 dB as opposed to 0.2 dB for an analytical model. The authors in [19] used a synthetic knowledge base (KB) built from the GNPy simulation tool, with three different ML techniques (RF, NN and KNN) to predict the generalized signal-to-noise ratio (GSNR) of unestablished lightpaths. The ANN-based QoT estimator was proven to produce the best results for the replicated European (EU) and USA network topologies with a mean absolute error (MAE) of 0.001 dB and 0.005 dB respectively.…”

Section: Neural Network For Lightpath Qot Estimationmentioning

confidence: 99%

See 2 more Smart Citations

Machine-Learning-Based Lightpath QoT Estimation and Forecasting

Allogba

Aladin

Tremblay

2022

J. Lightwave Technol.

View full text Add to dashboard Cite

Machine learning (ML) is more and more used to address the challenges of managing the physical layer of increasingly heterogeneous and complex optical networks. In this tutorial, we illustrate how simple and more sophisticated machine learning methods can be used in lightpath quality of transmission (QoT) estimation and forecast tasks. We also discuss data processing strategies with the aim to determine relevant features to feed the ML classifiers and predictors. We then introduce a preliminary study on the application of transfer learning to try to overcome the scarcity of field data.

show abstract