Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group
Journal Home About Issues in Progress Current Issue All Issues Early Posting Feature Issues

Reflective microring-resonator-based microwave photonic sensor incorporating a self-attention assisted convolutional neural network

Yeming Chen, Xiaoyi Tian, Joel Sved, Liwei Li, Luping Zhou, Linh Nguyen, and Xiaoke Yi

Open Access Open Access

Abstract

In this paper, a reflective microring resonator (MRR)-based microwave photonic (MWP) sensor incorporating a self-attention convolutional neural network (CNN) is presented. An MRR cascaded with an inverse-designed optical reflector is adopted as the sensor probe to allow for utilizing the responses generated from both the clockwise and counterclockwise resonant modes. Through the MWP interrogation, the cascaded resonant modes can be transformed into distinctive deep radio-frequency (RF) spectral notches under different modulator bias conditions. By using a self-attention assisted CNN processing to leverage both the local and global features of the RF spectra, a sensing model with improved accuracy can be established. As a proof of concept, the proposed scheme is experimentally demonstrated in temperature sensing. Even with a small dataset, the root-mean-square error of the sensing model established after training is achieved at 0.026°C, which shows a 10-fold improvement in sensing accuracy compared to that of the traditional linear fitting model.

© 2024 Optica Publishing Group

Full Article  |  PDF Article
More Like This
Convolutional neural network for improved event-based Shack-Hartmann wavefront reconstruction

Mitchell Grose, Jason D. Schmidt, and Keigo Hirakawa
Appl. Opt. 63(16) E35-E47 (2024)

PMONN: an optical neural network for photonic integrated circuits based on micro-resonator

Jingya Ding, Lianqing Zhu, Mingxin Yu, Lidan Lu, and Penghao Hu
Opt. Express 32(5) 7832-7847 (2024)

Microwave photonics frequency measurement with improved accuracy based on an artificial neural network

Xin An, Zhangyi Yang, Zuoheng Liu, Youdi Zhang, and Wei Dong
Appl. Opt. 63(10) 2535-2542 (2024)

View More...

Data availability

Data underlying the results presented in this paper are not publicly available at this time but maybe obtained from the authors upon reasonable request.

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (7)
Fig. 1.
Fig. 1. Schematic diagram of the proposed reflective MRR-based MWP sensing scheme assisted by deep learning. PD: photodetector.

Download Full Size | PDF

Fig. 2.
Fig. 2. (a) Schematic diagram of density-based optimization of the optical reflector. (b) Reflection spectra and S11-phase over 200 nm optical bandwidth. Inset: material geometry profile with optimized structure contour. (c) Simulated S11, $\angle {\rm S11}$ of structure with ${\rm Air}/{{\rm SiO}_2}$ cladding over 200 nm optical bandwidth.

Download Full Size | PDF

Fig. 3.
Fig. 3. Flow diagram of the convolutional self-attention model adopted to establish the regression model using the RF spectra acquired with three different DDMZM bias conditions.

Download Full Size | PDF

Fig. 4.
Fig. 4. (a) SEM of the inverse designed reflector. (b) Measured optical reflection spectrum of the inverse-designed reflector after calibration.

Download Full Size | PDF

Fig. 5.
Fig. 5. (a) Scanning electron micrograph of the microring with the inverse-designed reflector. (b) Comparison of the optical transmissions and phase change of the reflective MRR and standard MRR.

Download Full Size | PDF

Fig. 6.
Fig. 6. Measured three deep RF transmission notches at different DDMZM DC bias voltages. First optimal bias: $\pi$ phase shift; second optimal bias: ${3}\pi$ phase shift; third optimal bias: ${2}\pi$ phase shift.

Download Full Size | PDF

Fig. 7.
Fig. 7. (a) Notch frequency locations of the first RF spectrum in each measurement and their linear fitting results. (b) Estimated temperatures by the established convolutional self-attention model against the ground truth values. (c) Comparison of the RMSE performance of the convolutional neural network and the convolutional self-attention neural network. (d) Comparison of the RMSE performance of the convolutional self-attention model established with four different datasets.

Download Full Size | PDF

Equations (7)

Equations on this page are rendered with MathJax. Learn more.

H ( λ ) = a − r e − i 2 π λ n e f f L 1 − r a e i 2 π λ n e f f L e i ( π + 2 π λ n e f f L ) ,
H R ( λ ) = | H ( λ ) | 2 e i ( π − 2 ∡ H ( λ ) ) ,
E R = 4 log 10 ( a + r ) ( 1 − a r ) ( a − r ) ( 1 + a r ) .
Δ λ r e s = Δ n e f f L m ,
P R F ( f ) = η { γ | H R ( f C − f ) | 2 + 1 + 2 γ | H R ( f c − f ) | 2 × cos ⁡ ( Δ φ + ∡ H R ( f c − f ) ) } ,
γ | H R ( f C − f ) | 2 = 1 ,
cos ⁡ ( Δ φ + ∡ H R ( f c − f ) = − 1.
Select as filters
Select Topics Cancel
Top
       
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.
Privacy | Terms of Use