EUVIMEDEUVIMED entdecken
Crossref

Feedforward Neural Network Equalizers for SOA Nonlinearities in Coherent Optical Systems

Hamza Imtiaz, Arman Safarnejadian, Leslie A. Rusch, Ming Zeng

IEEE Photonics Technology Letters · 01.11.2026 · 38 (21) · 1475-1478

DOI: 10.1109/lpt.2026.3697278 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Photonics Technology Letters, erschienen am 01.11.2026. Verfasst von Hamza Imtiaz und 3 weiteren Autor:innen. Die Arbeit nennt 13 Literaturangaben. Gefördert von: Canada Natural Sciences and Engineering Research Council, Canada Foundation for Innovation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Semiconductor optical amplifiers (SOAs) operated as booster amplifiers provide a compact, power-efficient means to increase launch power and extend reach. However, SOA nonlinear gain dynamics introduce amplitude and phase distortions that degrade overall link performance. Post-compensation methods based on back-propagation have been explored for SOA nonlinearity mitigation in coherent links, but they rely on knowledge of SOA parameters that are difficult to estimate with sufficient accuracy. We examine a simple, data-driven feed-forward neural network (FFNN) equalizer to post-compensate SOA nonlinearities that does not require information on device parameters. The FFNN equalizer delivers Q-factor improvement of 3.4 dB and 2.5 dB over linear equalization for 16-quadrature amplitude modulation (QAM) signals at 30 Gbaud and 60 Gbaud, respectively. It maintains robust performance across two baud rates. We present a performance–complexity analysis that highlights key trade-offs and informs the choice of FFNN architecture.

Förderung

  • Canada Natural Sciences and Engineering Research Council
  • Canada Foundation for Innovation

Zitieren

APA

Imtiaz, H., Safarnejadian, A., Rusch, L. A., & Zeng, M. (2026). Feedforward Neural Network Equalizers for SOA Nonlinearities in Coherent Optical Systems. IEEE Photonics Technology Letters, 38(21), 1475-1478. https://doi.org/10.1109/lpt.2026.3697278

Vancouver

Imtiaz H, Safarnejadian A, Rusch LA, Zeng M. Feedforward Neural Network Equalizers for SOA Nonlinearities in Coherent Optical Systems. IEEE Photonics Technology Letters. 2026;38(21):1475-1478. doi:10.1109/lpt.2026.3697278

BibTeX
@article{Imtiaz2026,
  author = {Imtiaz, Hamza and Safarnejadian, Arman and Rusch, Leslie A. and Zeng, Ming},
  title = {Feedforward Neural Network Equalizers for SOA Nonlinearities in Coherent Optical Systems},
  journal = {IEEE Photonics Technology Letters},
  year = {2026},
  volume = {38},
  number = {21},
  pages = {1475--1478},
  doi = {10.1109/lpt.2026.3697278}
}
Kostenlos starten →