Nonlocal Scatterer for Compact Wave-Based Analog Computing

Goh, Heedong; Alù, Andrea

doi:10.1103/physrevlett.128.073201

Cited by 24 publications

(10 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…No general guideline exists for setting the ideal search length, for example, other than optimizing it after every (few) iterations or using more advanced algorithms . Second-order methods, such as the inexact Newton method, can be chosen as alternative solution methods to the steepest ascent or nonlinear conjugate gradient methods, which provide better search directions as well as optimal search lengths for highly nonlinear FOMs. , Another improvement to the implementation may be to use a nonlinear interpolation of the permittivity as a function of the density ς, as the linear interpolation we have used here can occasionally lead to instabilities and convergence problems . When it comes to fabrication constraints, more advanced methods than simply applying a Gaussian filter have also been implemented, ranging from a filter integrated in the control problem similarly to the binarization step we included to advanced methods incorporating constraints directly into the adjoint problem itself .…”

Section: Discussionmentioning

confidence: 99%

“…Next, we consider the constraint in the design variable, i.e., ς ∈ [0, 1]. To arrive at a smooth and unconstrained problem, we reparameterize ς ( r ) : double-struckR 3 → [ 0 , 1 ] ⊂ double-struckR with a new variable s ( r ) : double-struckR 3 → double-struckR , i.e., ς ( r ) = 1 2 [ 1 + tanh ⁡ s false( boldr false) ν ] Note that ς → 0 as s → −∞ and ς → 1 as s → ∞. The parameter ν controls the smoothness of the transition between the two bounds 1 and 0, where a smaller ν gives a sharper transition.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Inverse Design of Nonlinear Polaritonic Metasurfaces for Second Harmonic Generation

2023

Self Cite

View full text Add to dashboard Cite

Enabling strong nonlinear optical phenomena requires carefully designed photonic devices to maximize light–matter interactions. In recent years, topology optimization has become an increasingly popular tool to optimize photonic devices, due to its efficiency in dealing with extremely large parameter spaces. The focus has largely been on two-dimensional linear optical devices, particularly in integrated photonic settings, with a limited number of studies extending topology optimization to nonlinear effects in dielectric structures. Here, we apply topology optimization to metal–multiple quantum well–metal structures supporting intersubband transitions. In past years, the combination of photonic engineering and material design has enabled extremely efficient ultrathin second harmonic generating metasurfaces, based on L- and T-shaped resonators. Topology optimization enables novel designs that exceed the state-of-the-art nonlinear metasurfaces to date, with the promise of further improvement. The same approach may be extended to other frequency mixing phenomena, such as sum-frequency or difference-frequency generation.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Inverse Design of Nonlinear Polaritonic Metasurfaces for Second Harmonic Generation

2023

Self Cite

View full text Add to dashboard Cite

show abstract

“…Furthermore, Goh et al 63 . proved that any arbitrary shaped scatterer can be exploited by designing a meta-atom with an optimized nonlocal response using inverse design.…”

Section: All-optical Calculations At the Speed Of Lightmentioning

confidence: 99%

Computation at the speed of light: metamaterials for all-optical calculations and neural networks

2022

View full text Add to dashboard Cite

The explosion in the amount of information that is being processed is prompting the need for new computing systems beyond existing electronic computers. Photonic computing is emerging as an attractive alternative due to performing calculations at the speed of light, the change for massive parallelism, and also extremely low energy consumption. We review the physical implementation of basic optical calculations, such as differentiation and integration, using metamaterials, and introduce the realization of all-optical artificial neural networks. We start with concise introductions of the mathematical principles behind such optical computation methods and present the advantages, current problems that need to be overcome, and the potential future directions in the field. We expect that our review will be useful for both novice and experienced researchers in the field of all-optical computing platforms using metamaterials.

show abstract

“…In this work, a planar photonic chip composed of an all-dielectric multilayer structure is proposed and its nonlocality (or named as angular-dependent transmission) is tailored to enable signal manipulation in the momentum domain 28 , 29 . It can be used to perform various types of mathematical operations on optical signals, such as the first-, second-, third-, fourth- and even higher-order spatial derivative operations without altering the chip’s structural parameters.…”

Section: Introductionmentioning

confidence: 99%

Single planar photonic chip with tailored angular transmission for multiple-order analog spatial differentiator

et al. 2022

View full text Add to dashboard Cite

Analog spatial differentiation is used to realize edge-based enhancement, which plays an important role in data compression, microscopy, and computer vision applications. Here, a planar chip made from dielectric multilayers is proposed to operate as both first- and second-order spatial differentiator without any need to change the structural parameters. Third- and fourth-order differentiations that have never been realized before, are also experimentally demonstrated with this chip. A theoretical analysis is proposed to explain the experimental results, which furtherly reveals that more differentiations can be achieved. Taking advantages of its differentiation capability, when this chip is incorporated into conventional imaging systems as a substrate, it enhances the edges of features in optical amplitude and phase images, thus expanding the functions of standard microscopes. This planar chip offers the advantages of a thin form factor and a multifunctional wave-based analogue computing ability, which will bring opportunities in optical imaging and computing.

show abstract

Nonlocal Scatterer for Compact Wave-Based Analog Computing

Cited by 24 publications

References 34 publications

Inverse Design of Nonlinear Polaritonic Metasurfaces for Second Harmonic Generation

Inverse Design of Nonlinear Polaritonic Metasurfaces for Second Harmonic Generation

Computation at the speed of light: metamaterials for all-optical calculations and neural networks

Single planar photonic chip with tailored angular transmission for multiple-order analog spatial differentiator

Contact Info

Product

Resources

About