Micro ring resonators as building blocks for an all-optical high-speed reservoir-computing bit-pattern-recognition system

Mesaritakis, Charis; Papataxiarhis, Vassilis; Syvridis, Dimitris

doi:10.1364/josab.30.003048

Cited by 85 publications

(42 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In a previous paper, we investigated the most important properties of photonic reservoirs by means of simulations for an isolated spoken digit recognition task 19 . The reservoir employed there was a network of coupled semiconductor optical amplifiers.…”

Section: Resultsmentioning

confidence: 99%

“…The model used to describe the optical behaviour of the chip consists of a time-domain approach where a complex-valued envelope E(t) of the signal with carrier frequency 2pc l is propagated through a passive network. Taking into account effects like a finite propagation time t, combined with a waveguide loss a and phase change of DF, the transmission through the interconnections is modelled by 19 E out ðtÞ ¼ exp À a 2 À Á expð À jD FÞE in ðt À tÞ. Splitters and combiners are considered to be lossless and are simply modelled by division of the power by a factor 2 and complex in-phase addition, respectively.…”

Section: Methodsmentioning

confidence: 99%

“…This is particularly attractive when the information is already in the optical domain as in the case of many telecom and image processing applications. Optical reservoirs based on a fibre and one dynamical node [14][15][16][17][18] as well as reservoirs based on ring resonators 19 have been demonstrated. In our own previous work we have shown through reservoir simulations that integrated optical chips with a network of coupled semiconductor optical amplifiers can also be used, with the advantage of a much smaller footprint 20 .…”

mentioning

confidence: 99%

“…Here, we propose an alternative architecture that circumvents these problems completely. As a reservoir, we use a passive silicon photonics chip that only contains waveguides, splitters and combiners, that is, we eliminate the amplifiers we used in our previous work 19 . As a consequence, the required nonlinearity is no longer present in the reservoir itself, and the signals in the output waveguides of each node are linear superpositions of the complex amplitudes of the input waveguides of that node.…”

mentioning

confidence: 99%

See 3 more Smart Citations

Experimental demonstration of reservoir computing on a silicon photonics chip

et al. 2014

View full text Add to dashboard Cite

In today's age, companies employ machine learning to extract information from large quantities of data. One of those techniques, reservoir computing (RC), is a decade old and has achieved state-of-the-art performance for processing sequential data. Dedicated hardware realizations of RC could enable speed gains and power savings. Here we propose the first integrated passive silicon photonics reservoir. We demonstrate experimentally and through simulations that, thanks to the RC paradigm, this generic chip can be used to perform arbitrary Boolean logic operations with memory as well as 5-bit header recognition up to 12.5 Gbit s À 1 , without power consumption in the reservoir. It can also perform isolated spoken digit recognition. Our realization exploits optical phase for computing. It is scalable to larger networks and much higher bitrates, up to speeds 4100 Gbit s À 1 . These results pave the way for the application of integrated photonic RC for a wide range of applications.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Experimental demonstration of reservoir computing on a silicon photonics chip

et al. 2014

View full text Add to dashboard Cite

show abstract

“…The different streams of bits are separated at the output port without the need for deconvolution processes (such as Fourier transforms). Furthermore in the last years scientific community started to recognize the potential of photonics to achieve parallel (vectorial) computation [4][5][6][7] and to define logical paradigms beyond the binary one, able to perform complex function within a single operation [8,9].…”

Section: Introductionmentioning

confidence: 99%

Hybrid Materials for Integrated Photonics

Bettotti

2014

Advances in Optics

View full text Add to dashboard Cite

In this review materials and technologies of the hybrid approach to integrated photonics (IP) are addressed. IP is nowadays a mature technology and is the most promising candidate to overcome the main limitations that electronics is facing due to the extreme level of integration it has achieved. IP will be based on silicon photonics in order to exploit the CMOS compatibility and the large infrastructures already available for the fabrication of devices. But silicon has severe limits especially concerning the development of active photonics: its low efficiency in photons emission and the limited capability to be used as modulator require finding suitable materials able to fulfill these fundamental tasks. Furthermore there is the need to define standardized processes to render these materials compatible with the CMOS process and to fully exploit their capabilities. This review describes the most promising materials and technological approaches that are either currently implemented or may be used in the coming future to develop next generations of hybrid IP devices.

show abstract

Emergent Self‐Adaptation in an Integrated Photonic Neural Network for Backpropagation‐Free Learning

Lugnan,

Aggarwal,

Brückerhoff‐Plückelmann

et al. 2024

Advanced Science

View full text Add to dashboard Cite

Plastic self‐adaptation, nonlinear recurrent dynamics and multi‐scale memory are desired features in hardware implementations of neural networks, because they enable them to learn, adapt, and process information similarly to the way biological brains do. In this work, these properties occurring in arrays of photonic neurons are experimentally demonstrated. Importantly, this is realized autonomously in an emergent fashion, without the need for an external controller setting weights and without explicit feedback of a global reward signal. Using a hierarchy of such arrays coupled to a backpropagation‐free training algorithm based on simple logistic regression, a performance of 98.2% is achieved on the MNIST task, a popular benchmark task looking at classification of written digits. The plastic nodes consist of silicon photonics microring resonators covered by a patch of phase‐change material that implements nonvolatile memory. The system is compact, robust, and straightforward to scale up through the use of multiple wavelengths. Moreover, it constitutes a unique platform to test and efficiently implement biologically plausible learning schemes at a high processing speed.

show abstract

Micro ring resonators as building blocks for an all-optical high-speed reservoir-computing bit-pattern-recognition system

Cited by 85 publications

References 26 publications

Experimental demonstration of reservoir computing on a silicon photonics chip

Experimental demonstration of reservoir computing on a silicon photonics chip

Hybrid Materials for Integrated Photonics

Emergent Self‐Adaptation in an Integrated Photonic Neural Network for Backpropagation‐Free Learning

Contact Info

Product

Resources

About