EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Neural-network-assisted polarization-robust FBG interrogation based on an integrated SOI-AWG

Qichao Wang, Jiawen Lu, Yifei Chen, Dongyu Wang, Wanghua Zhu, Enze Zhou, Binfeng Yun, Guohua Hu, Yiping Cui

APL Photonics · 01.07.2026 · 11 (7)

DOI: 10.1063/5.0336489 ↗

Kurz gefasst

Fachartikel in der Zeitschrift APL Photonics, online erschienen am 07.07.2026. Verfasst von Qichao Wang und 8 weiteren Autor:innen. Die Arbeit nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Silicon-on-insulator (SOI)-based on-chip interrogation is a promising approach for realizing compact, low-power, and low-cost fiber Bragg grating sensing systems. However, SOI-based arrayed waveguide grating (AWG) interrogators often suffer from a trade-off between a wide continuous interrogation range and wavelength accuracy. In addition, the high refractive index contrast of the SOI platform introduces spectral distortions and strong polarization sensitivity, which complicate accurate wavelength interrogation across a broad bandwidth. Here, a polarization-robust on-chip wavelength interrogation chip based on a cascaded polarization beam splitter (PBS) and AWG is proposed and fabricated on the SOI platform, along with a lightweight neural-network-assisted readout. The integrated PBS stabilizes the polarization state entering the AWG, while a wide-overlap AWG design enables continuous interrogation across adjacent channels. A feature-engineered feedforward neural network directly maps multi-channel spectral responses to wavelength without requiring cascaded models or spectral segmentation. Experimental results demonstrate high-accuracy interrogation with a root-mean-square error of 1.53 pm over the range of 1557.39–1568.85 nm. Stable picometer-level accuracy is maintained even under significant polarization-induced signal variations.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Wang, Q., Lu, J., Chen, Y., Wang, D., Zhu, W., Zhou, E., Yun, B., Hu, G., & Cui, Y. (2026). Neural-network-assisted polarization-robust FBG interrogation based on an integrated SOI-AWG. APL Photonics, 11(7). https://doi.org/10.1063/5.0336489

Vancouver

Wang Q, Lu J, Chen Y, Wang D, Zhu W, Zhou E, et al. Neural-network-assisted polarization-robust FBG interrogation based on an integrated SOI-AWG. APL Photonics. 2026;11(7). doi:10.1063/5.0336489

BibTeX
@article{Wang2026,
  author = {Wang, Qichao and Lu, Jiawen and Chen, Yifei and Wang, Dongyu and Zhu, Wanghua and Zhou, Enze and Yun, Binfeng and Hu, Guohua and Cui, Yiping},
  title = {Neural-network-assisted polarization-robust FBG interrogation based on an integrated SOI-AWG},
  journal = {APL Photonics},
  year = {2026},
  volume = {11},
  number = {7},
  doi = {10.1063/5.0336489}
}
Kostenlos starten →