Simulating self-learning in photorefractive optical reservoir computers

Laporte, Floris; Dambre, Joni; Bienstman, Peter

doi:10.1038/s41598-021-81899-w

Cited by 10 publications

(4 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To make the FMD and adjoint method presented in this work more accessible, we have released an opensource FDTD and FDFD package that features gradients computed by all three methods outlined in this paper [22]. Our implementation makes use of automatic differentiation to provide flexible usage and more robust computation [23][24][25] V. CONCLUSION…”

Section: Discussionmentioning

confidence: 99%

Forward-Mode Differentiation of Maxwell’s Equations

et al. 2019

View full text Add to dashboard Cite

We present a previously unexplored forward-mode differentiation method for Maxwell's equations, with applications in the field of sensitivity analysis. This approach yields exact gradients and is similar to the popular adjoint variable method, but provides a significant improvement in both memory and speed scaling for problems involving several output parameters, as we analyze in the context of finite-difference time-domain (FDTD) simulations. Furthermore, it provides an exact alternative to numerical derivative methods, based on finite-difference approximations. To demonstrate the usefulness of the method, we perform sensitivity analysis of two problems. First we compute how the spatial near-field intensity distribution of a scatterer changes with respect to its dielectric constant. Then, we compute how the spectral power and coupling efficiency of a surface grating coupler changes with respect to its fill factor.

show abstract

Section: Discussionmentioning

confidence: 99%

Forward-Mode Differentiation of Maxwell’s Equations

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Chaotic results generated from multigrating phase conjugation in a photorefractive four-wave mixing process has also been reported [35]. The rich and complex spatio-temporal nonlinear dynamics exhibited by the photorefractive process makes it a promising candidate for artificial intelligence or neural network-based applications, such as reservoir computing [36,37].…”

Section: Introductionmentioning

confidence: 93%

Period-Doubling Route to Chaos in Photorefractive Two-Wave Mixing

Saju,

Kinashi,

Tsutsumi

et al. 2024

Photonics

View full text Add to dashboard Cite

This paper investigates the possibilities of complex nonlinear dynamic signal generation using a simple photorefractive two-wave mixing system without any feedback using numerical simulations. The novel idea is to apply a sinusoidal electric field to the system inroder to extract nonlinear dynamic behavior. The mathematical model of the system was constructed using Kogelnick’s coupled wave equations and Kukhtarev’s material equation. The spatio-temporal evolution of the system was simulated in discrete steps numerically. The temporal evolution of the output light intensity exhibits period doubling, behavior which is a characteristic feature of complex nonlinear dynamic systems. A bifurcation diagram and Lyapunov exponentials confirm the presence of the period-doubling route to chaos in the system. The observed complex signal pattern varies uniformly with respect to the amplitude of the applied field, indicating a controllable nonlinear dynamic behavior. Such a system can be very useful in applications such as photonic reservoir computing, in-materio computing, photonic neuromorphic networks, complex signal detection, and time series prediction.

show abstract

“…The methods of dynamic holography are among the promising ones for transforming the parameters of laser beams and images. They cover many applications: the control over laser beam parameters [5,6]; the development of various sensors [7][8][9]; optical phase conjugation (OPC), which is used to create mirrors with an inverted wavefront, which is very necessary for lidar systems and photolithography [10]; the development of systems for optical computing; and the creation of optical computers and holographic artificial intelligence [11,12].…”

Section: Introductionmentioning

confidence: 99%

Керування Інтенсивністю Лазерного Випромінювання У Рідкокристалічних Вентилях При Записі Об’ємної Динамічної Ґратки

Mystetskyi,

Bugaychuk

2023

Ukr. J. Phys.

View full text Add to dashboard Cite

Експериментальнi дослiдження номiнально чистих нематичних рiдких кристалiв (НРК) пiдтверджують запис динамiчних голографiчних ґраток у комiрках як з гомеотропною орiєнтацiєю, так i з планарною. Пояснення можна знайти, виходячи iз фоторефрактивного механiзму запису ґратки, особливiстю якого являється формування нерiвноважного заряду на поверхнi пiдкладинки комiрки пiд дiєю просторово неоднорiдного свiтлового поля. Поява внутрiшнього тангенцiального електричного поля (вздовж пiдкладинок комiрки), разом iз зовнiшнiм електричним полем, що прикладається нормально до пiдкладинок комiрки, вiдкриває додатковi можливостi у керуваннi напрямку вектора результуючого електричного поля. В данiй роботi розроблена i аналiзується модель змiни iнтенсивностей лазерних променiв при їх самодифракцiї i дифракцiї на динамiчнiй ґратцi, створенiй в НРК. Динамiчна фазова ґратка формується завдяки орiєнтацiйному механiзму двозаломлення в НРК при двопучковiй взаємодiї лазерних променiв, що утворюють просторово перiодичну iнтерференцiйну картину дiючого свiтлового поля. Результати проведених розрахункiв вихiдних iнтенсивностей лазерних променiв в перших порядках самодифракцiї i дифракцiї добре узгоджуються з експериментальними вимiрюваннями. Зокрема, вони пояснюють залежнiсть дифракцiйної ефективностi вiд величини зовнiшньої прикладеної напруги, що має добре виражений максимум.

show abstract

Simulating self-learning in photorefractive optical reservoir computers

Cited by 10 publications

References 42 publications

Forward-Mode Differentiation of Maxwell’s Equations

Forward-Mode Differentiation of Maxwell’s Equations

Period-Doubling Route to Chaos in Photorefractive Two-Wave Mixing

Керування Інтенсивністю Лазерного Випромінювання У Рідкокристалічних Вентилях При Записі Об’ємної Динамічної Ґратки

Contact Info

Product

Resources

About