(47-4) 10 * << * >> * Russian * English * Content * All Issues

On chip optical neural networks based on MMI microring resonators for image classification
T.T. Bui 1, D.T. Le 2, T.H.L. Nguyen 3, T.T. Le 2

1 FPT University Hoa Lac High Tech Park, 100000, Hanoi, Viet Nam;
2 International School (VNU-IS), Vietnam National University (VNU)-Hanoi,
100000, Hanoi, Viet Nam, Xuan Thuy St. 144;
3 Hanoi University of Natural Resources and Environment,
100000, Hanoi, Viet Nam, Phu Dien St. 41A

 PDF, 2823 kB

DOI: 10.18287/2412-6179-CO-1211

Pages: 588-595.

Full text of article: English language.

Abstract:
We propose a new on-chip optical neural network (OONN) based on multimode interference-microring resonators (MMI-RRs). The suggested structure eliminates the need for wavelength division multiplexers (WDM) to create an optical neuron on a single chip. New microring resonator structure based on 4×4 MMI coupler with a size of 24µm × 2900 µm is used for the basic elements of the computation matrix, as a result a higher bandwidth and free spectral range (FSR) can be achieved. The Si3N4 platform along with the graphene sheet is designed to modulate the signals and weights of the neural networks at a very high speed. The Si3N4 can provide wide range of operating wavelengths and can work directly with the wavelengths of color images. The structure's benefits include rapid computing speed, little loss, and the ability to handle both positive and negative values. The OONN has been applied to the MNIST dataset with a speed faster than 2.8 to 14x times compared with the conventional GPU methods.

Keywords:
all-optical dot product, image processing, multimode interference coupler, optical convolutional neural networks, optical signal processing, microring resonators, silicon photonics.

Citation:
Bui TT, Le DT, Nguyen THL and Le TT. On chip optical neural networks based on MMI microring resonators for image classification. Computer Optics 2023; 47(4): 588-595. DOI: 10.18287/2412-6179-CO-1211.

Acknowledgements:
This research is funded by Vietnam National Founda-tion for Science and Technology Development (NA-FOSTED) under grant number 103.03-2018.354.

References:
  1. Xiang S, et al. A review: Photonics devices, architectures, and algorithms for optical neural computing. J Semicond 2021; 42(2): 023105. DOI: 10.1088/1674-4926/42/2/023105.
  2. Kazanskiy NL, Butt MA, Khonina SN. Optical computing: Status and perspectives. Nanomaterials 2022; 12(13): 2171. DOI: 10.3390/nano12132171.
  3. Sui X, Wu Q, Liu J, Chen Q, Gu G. A review of optical neural networks. IEEE Access 2020; 8: 70773-70783. DOI: 10.1109/ACCESS.2020.2987333.
  4. Yang L, Ji R, Zhang L, Ding J, Xu Q. On-chip CMOS-compatible optical signal processor. Opt Express 2012; 20(12): 13560-13565. DOI: 10.1364/OE.20.013560.
  5. Salmani M, Eshaghi A, Luan E, Saha S. Photonic computing to accelerate data processing in wireless communications. Opt Express 2021; 29(14): 22299-22314. DOI: 10.1364/OE.423747.
  6. Harris NC, et al. Linear programmable nanophotonic processors. Optica 2018; 5(12): 1623-1631. DOI: 10.1364/OPTICA.5.001623.
  7. Tait AN, et al. Feedback control for microring weight banks. Opt Express 2018; 26(20): 26422-26443. DOI: 10.1364/OE.26.026422.
  8. Le TT, Cahill LW, Elton D. The design of 2x2 SOI MMI couplers with arbitrary power coupling ratios. Electron Lett 2009; 45(22): 1118-1119.
  9. Ferreira de Lima T, et al. Design automation of photonic resonator weights. Nanophotonics 2022; 11(4-5): 49. DOI: 10.1515/nanoph-2022-0049.
  10. Zhang D, Tan Z. A review of optical neural networks. Appl Sci 2022; 12(11): 5338. DOI: 10.3390/app12115338.
  11. Tait AN, Nahmias MA, Shastri BJ, Prucnal PR. Broadcast and weight: An integrated network for scalable photonic spike processing. J Lightw Technol 2014; 32(21): 4029-4041. DOI: 10.1109/JLT.2014.2345652.
  12. Liu J, Khan ZU, Wang C, Zhang H, Sarjoghian S. Review of graphene modulators from the low to the high figure of merits. J Phys D: Appl Phys 2020; 53(23): 233002. DOI: 10.1088/1361-6463/ab7cf6.
  13. Xu X, et al. 11 TOPS photonic convolutional accelerator for optical neural networks. Nature 2021; 589(7840): 44-51. DOI: 10.1038/s41586-020-03063-0.
  14. Zhang H, et al. An optical neural chip for implementing complex-valued neural network. Nat Commun 2021; 12(1): 457. DOI: 10.1038/s41467-020-20719-7.
  15. Bachmann M, Besse PA, Melchior H. General self-imaging properties in N x N multimode interference couplers including phase relations. Appl Opt 1994; 33(18): 3905-3911.
  16. Le TT. Multimode interference structures for photonic signal processing. LAP Lambert Academic Publishing; 2010.
  17. Bao Q. 2D Materials for photonic and optoelectronic applications. Woodhead Publishing; 2019.
  18. Xing P, Ooi KJA, Tan DTH, "Ultra-broadband and compact graphene-on-silicon integrated waveguide mode filters. Sci Rep 2018; 8(1): 9874. DOI: 10.1038/s41598-018-28076-8.
  19. Hanson GW. Dyadic Green’s functions and guided surface waves for a surface conductivity model of graphene. J Appl Phys 2008; 103(6): 064302. DOI: 10.1063/1.2891452.
  20. Capmany J, Domenech D, Muoz P. Silicon graphene Bragg gratings. Opt Express 2014; 22(5): 5283-5290. DOI: 10.1364/OE.22.005283.
  21. Chremmos I, Schwelb O. Photonic microresonator research and applications. New York: Springer Science+Business Media LLC; 2010.
  22. Rumley S, et al. Optical interconnects for extreme scale computing systems. Parallel Comput 2017; 64: 65-80. DOI: 10.1016/j.parco.2017.02.001.
  23. Bangari V, et al. Digital electronics and analog photonics for convolutional neural networks (DEAP-CNNs). IEEE J Sel Top Quantum Electron 2020; 26(1): 5100209. DOI: 10.1109/JSTQE.2019.2945540.
  24. Zhang W, et al. Silicon microring synapses enable photonic deep learning beyond 9-bit precision. Optica 2022; 9(5): 579-584. DOI: 10.1364/OPTICA.446100.
  25. Wu L, Liu H, Li J, Wang S, Qu S, Dong L. A 130 GHz electro-optic ring modulator with double-layer grapheme. Crystals 2017; 7(3): 65. DOI: 10.3390/cryst7030065.
  26. AMD Radeon™ Instinct™ MI25 Accelerator. 2022. Source: <https://www.amd.com/en/products/professional-graphics/instinct-mi25>.
  27. NVidia. GeForce. Specifications. GeForce GTX 1080 Ti2022. Source: <https://www.nvidia.com/en-gb/geforce/graphics-cards/geforce-gtx-1080-ti/specifications/>.

© 2009, IPSI RAS
151, Molodogvardeiskaya str., Samara, 443001, Russia; E-mail: journal@computeroptics.ru ; Tel: +7 (846) 242-41-24 (Executive secretary), +7 (846) 332-56-22 (Issuing editor), Fax: +7 (846) 332-56-20