Journals Articles Conferences News About CLP
Submit a paper Email Alert
Search
Search
Articles
Journals
News
Advanced Search
Top Searches
2D MaterialsTransformation opticsQuantum PhotonicsSmall lasersSemiconductor UV PhotonicsSupersymmetric microring laser arrays

Photonics Research, Volume. 11, Issue 12, 2033(2023)

Spiking information processing in a single photonic spiking neuron chip with double integrated electronic dendrites

Yahui Zhang1,2, Shuiying Xiang1,2、*, Xingxing Guo1, Yanan Han1, Yuechun Shi3, Xiangfei Chen4, Genquan Han2, and Yue Hao2
Author Affiliations
  • 1State Key Laboratory of Integrated Service Networks, Xidian University, Xi’an 710071, China
  • 2State Key Discipline Laboratory of Wide Bandgap Semiconductor Technology, Xidian University, Xi’an 710071, China
  • 3Yongjiang Laboratory, Ningbo 315202, China
  • 4Key Laboratory of Intelligent Optical Sensing and Manipulation, Ministry of Education, National Laboratory of Solid State Microstructures, College of Engineering and Applied Sciences, Institute of Optical Communication Engineering, Nanjing University, Nanjing 210023, China
    • show less
    Full TextGet PDF
    Figures&Tables (8)
    Equations (6)
    References (44)
    Cited By (3)
    Tools
    Paper Information
    Topics
    Figures & Tables(8)
    Operation principle, chip, and experimental setup of FP-SA neuron. (a) Illustration of a biological neuron model receiving excitatory and inhibitory stimulations at functional dendrites. (b) Microscope image of the FP-SA chip and the operation principle of the FP-SA neuron. (c) Experimental setup of the encoding system based on a single FP-SA neuron with double electronic dendrites. Source, voltage source and current source; FPGA, field programmable gate array; CH1, CH2, two channels of FPGA; Bias Tee 1, Bias Tee 2, bias tees; FP-SA, integrated Fabry–Perot laser with an intracavity saturable absorber; OI, optical isolator; VOA, variable optical attenuator; OC, optical coupler; PD, photodetector; OSC, oscilloscope; OSA, optical spectrum analyzer. Red line, electrical signal; blue line, optical signal.

    Fig. 1. Operation principle, chip, and experimental setup of FP-SA neuron. (a) Illustration of a biological neuron model receiving excitatory and inhibitory stimulations at functional dendrites. (b) Microscope image of the FP-SA chip and the operation principle of the FP-SA neuron. (c) Experimental setup of the encoding system based on a single FP-SA neuron with double electronic dendrites. Source, voltage source and current source; FPGA, field programmable gate array; CH1, CH2, two channels of FPGA; Bias Tee 1, Bias Tee 2, bias tees; FP-SA, integrated Fabry–Perot laser with an intracavity saturable absorber; OI, optical isolator; VOA, variable optical attenuator; OC, optical coupler; PD, photodetector; OSC, oscilloscope; OSA, optical spectrum analyzer. Red line, electrical signal; blue line, optical signal.

    Download full size
    View in Article
    Optical spectra and spiking dynamics of the FP-SA neuron. (a) Optical spectra of the FP-SA neuron. (a1) Optical spectrum with IG=50 mA, VSA=0 V; (a2) optical spectrum with IG=91 mA, VSA=−3.8 V. (b) Spikes generated by the FP-SA neuron. (b1) IG=91 mA, VSA=−3.6 V; (b2) IG=100 mA, VSA=−3.6 V; (b3) IG=100 mA, VSA=−3.8 V. (c), (d) Frequencies as functions of IG and VSA.

    Fig. 2. Optical spectra and spiking dynamics of the FP-SA neuron. (a) Optical spectra of the FP-SA neuron. (a1) Optical spectrum with IG=50  mA, VSA=0  V; (a2) optical spectrum with IG=91  mA, VSA=−3.8  V. (b) Spikes generated by the FP-SA neuron. (b1) IG=91  mA, VSA=−3.6  V; (b2) IG=100  mA, VSA=−3.6  V; (b3) IG=100  mA, VSA=−3.8  V. (c), (d) Frequencies as functions of IG and VSA.

    Download full size
    View in Article
    Frequency encoding in FP-SA neuron. (a1)–(a6) Excitatory stimulus and corresponding response in FP-SA with IG=90 mA and VSA=−3.9 V. (b1)–(b6) Inhibitory stimulus and corresponding response in FP-SA with IG=90 mA and VSA=−3.9 V. (c1)–(c5) Simultaneous excitatory and inhibitory stimuli and corresponding response in FP-SA with IG=81 mA and VSA=−3.9 V. (a1)–(c1) Excitatory stimulus for gain section. (a2)–(c2) Inhibitory stimulus for saturable absorber section. (a3)–(b3) Response of FP-SA neuron. (a4)–(b4) Spike time interval of spikes in (a3)–(c3). (a5)–(b5) Frequency of spikes in (a3)–(c3). (a6)–(c6) Eye diagram of (a5)–(c5).

    Fig. 3. Frequency encoding in FP-SA neuron. (a1)–(a6) Excitatory stimulus and corresponding response in FP-SA with IG=90  mA and VSA=−3.9  V. (b1)–(b6) Inhibitory stimulus and corresponding response in FP-SA with IG=90  mA and VSA=−3.9  V. (c1)–(c5) Simultaneous excitatory and inhibitory stimuli and corresponding response in FP-SA with IG=81  mA and VSA=−3.9  V. (a1)–(c1) Excitatory stimulus for gain section. (a2)–(c2) Inhibitory stimulus for saturable absorber section. (a3)–(b3) Response of FP-SA neuron. (a4)–(b4) Spike time interval of spikes in (a3)–(c3). (a5)–(b5) Frequency of spikes in (a3)–(c3). (a6)–(c6) Eye diagram of (a5)–(c5).

    Download full size
    View in Article
    Spatiotemporal encoding in FP-SA neuron. (a1)–(a4) Excitatory stimulus and corresponding response in FP-SA with IG=65 mA and VSA=−4.6 V. (b1)–(b4) Inhibitory stimulus and corresponding response in FP-SA with IG=64 mA and VSA=−4.19 V. (c1)–(c4) Simultaneous excitatory and inhibitory stimuli and corresponding response in FP-SA with IG=61.6 mA and VSA=−3.97 V. (a1)–(c1) Excitatory stimulus for gain section. (a2)–(c2) Inhibitory stimulus for saturable absorber section. (a3)–(b3) Response of FP-SA neuron. (a4)–(b4) Overlapped color map of the responses of FP-SA neuron.

    Fig. 4. Spatiotemporal encoding in FP-SA neuron. (a1)–(a4) Excitatory stimulus and corresponding response in FP-SA with IG=65  mA and VSA=−4.6  V. (b1)–(b4) Inhibitory stimulus and corresponding response in FP-SA with IG=64  mA and VSA=−4.19  V. (c1)–(c4) Simultaneous excitatory and inhibitory stimuli and corresponding response in FP-SA with IG=61.6  mA and VSA=−3.97  V. (a1)–(c1) Excitatory stimulus for gain section. (a2)–(c2) Inhibitory stimulus for saturable absorber section. (a3)–(b3) Response of FP-SA neuron. (a4)–(b4) Overlapped color map of the responses of FP-SA neuron.

    Download full size
    View in Article
    Schematic diagram of photonic SNN for the recognition of Iris data set. (a1) Photonic SNN based on frequency encoding. (a2) Photonic SNN based on spatiotemporal encoding. (b) Detailed connection of presynaptic neuron and postsynaptic neuron.

    Fig. 5. Schematic diagram of photonic SNN for the recognition of Iris data set. (a1) Photonic SNN based on frequency encoding. (a2) Photonic SNN based on spatiotemporal encoding. (b) Detailed connection of presynaptic neuron and postsynaptic neuron.

    Download full size
    View in Article
    (a1), (a2) Recognition accuracy of the Iris data set and (b1)–(c2) weight distribution of corresponding photonic SNNs. (a1)–(c1) Corresponding to the architecture of Fig. 5(a1) with the frequency encoding; (a2)–(c2) corresponding to the architecture of Fig. 5(a2) with the spatiotemporal encoding. (b1), (b2) Weight range is [−1, 1]; (c1), (c2) weight rang is [0, 1].

    Fig. 6. (a1), (a2) Recognition accuracy of the Iris data set and (b1)–(c2) weight distribution of corresponding photonic SNNs. (a1)–(c1) Corresponding to the architecture of Fig. 5(a1) with the frequency encoding; (a2)–(c2) corresponding to the architecture of Fig. 5(a2) with the spatiotemporal encoding. (b1), (b2) Weight range is [−1, 1]; (c1), (c2) weight rang is [0, 1].

    Download full size
    View in Article
    Recognition accuracy of the Iris data set in noisy photonic SNNs. (a) Corresponding to the architecture of Fig. 5(a1) with the frequency encoding; (b) corresponding to the architecture of Fig. 5(a2) with the spatiotemporal encoding.

    Fig. 7. Recognition accuracy of the Iris data set in noisy photonic SNNs. (a) Corresponding to the architecture of Fig. 5(a1) with the frequency encoding; (b) corresponding to the architecture of Fig. 5(a2) with the spatiotemporal encoding.

    Download full size
    View in Article

    • Table 1. Two-Section Laser Parameters

      View table
      View in Article

      Table 1. Two-Section Laser Parameters

      Param.DescriptionValue
      VGGain section cavity volume2.4×10−18  m3
      VSASA section cavity volume2.4×10−18  m3
      ΓGGain section confinement factor0.06
      ΓSASA section confinement factor0.05
      τGGain section carrier lifetime1 ns
      τSASA section carrier lifetime100 ps
      τphPhoton lifetime4.8 ps
      gGGain section differential gain/loss2.9×10−12  m3 s−1
      gSASA section differential gain/loss14.5×10−12  m3 s−1
      n0GGain section transparency carrier density1.1×1024  m−3
      n0SASA section transparency carrier density0.89×1024  m−3
      BrBimolecular recombination term10×10−16  m3 s−1
      ηcOutput power coupling coefficient0.4
      βSpontaneous emission coupling factor1×10−4
      cSpeed of light3×108  m/s
      hPlanck constant6.63×10−34  J  s
      eElectron charge1.6×10−19  C
    Tools

    Get Citation

    Copy Citation Text

    Yahui Zhang, Shuiying Xiang, Xingxing Guo, Yanan Han, Yuechun Shi, Xiangfei Chen, Genquan Han, Yue Hao, "Spiking information processing in a single photonic spiking neuron chip with double integrated electronic dendrites," Photonics Res. 11, 2033 (2023)

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Set citation alerts for article
    Save article for my favorites
    Paper Information

    Category: Lasers and Laser Optics

    Received: Jul. 5, 2023

    Accepted: Sep. 29, 2023

    Published Online: Nov. 20, 2023

    The Author Email: Shuiying Xiang (jxxsy@126.com)

    DOI:10.1364/PRJ.499767

    Topics

    laser devices and laser physics
    Lasers and Laser Optics
    Laser physics
    laser manufacturing
    Instrumentation, Measurement and Metrology