Graphene/Al2O3/Graphene heterojunction for Light emitting tunneling diode

Demultiplexing OAM beams via Fourier optical convolutional neural network

“…141,142 The most straightforward and prevalent utilized approach is conjugate mode sorting demultiplexing technique, leveraging the orthogonal properties of the modes. [143][144][145] In this demultiplexing layout, the dot-product interaction with the conjugate modes yields a spatial light distribution featuring a pronounced central peak (with no discernible doughnut shape). The demultiplexing scheme involves equally dividing the power of the transmitted multiplexed data (carrying different OAMs) and configuring the SLM at each termination of the system's detecting arm to replicate the phase distribution of the corresponding propagating mode at the input.…”

Free-space optical multiplexed orbital angular momentum beam identification system using Fourier optical convolutional layer based on 4f system

“…After two decades development, kernels generated by state-in-the-art DMD outperforms the older hologram mask with a higher enough throughput to simulate completely neural planes, makes optical convolutional efficiency promising. [87][88][89][90][91][92][93][94][95][96][97][98][99][100][101] In this study, we experimentally generate multiplexed OAM beams and introduce optical filtering method which relys on spatial Fourier transform of images in the frequency domain as optical convolutional neural network to train and identify multiplexed OAM beams under simulated atmospheric turbulence conditions, we show the system currently capable of classifying 12 classes at test accuracy of 95% (under weak turbulence) and 87% (under strong turbulence).…”

Exploring the potential of high-speed 2D and 3D materials in silicon photonics

“…The photonic processors are applicable to a vast array of applications, including high-performance computing, data centers, artificial intelligence, quantum computing, highspeed communication system [1]- [3], GPU accelerators [4], and photonic computing [5]- [8]. The most critical part of the research of silicon photonics is the exploration of high-speed 2D and 3D materials and corresponding applications such as light sources [9], electric-optical modulators [10]- [23], and photodetectors. Two-dimensional materials [24]- [33] such as graphene and transition metal dichalcogenides have been widely used as gate insulators and channel materials for field-effect transistors and active tunable materials for the optical modulator and photodetectors due to their high conductivity and unique bandgap structure.…”

“…Another important property of graphene is that its absorption is independent of frequency over a broad operational band. This property makes graphene a promising material for use in optical communication systems, where high-speed broadband modulation is necessary [9], [10], [24], [44], [55]- [75]. parameters, and the potential complementary-metal-oxide-semiconductor (CMOS) compatibility.…”