Towards integrated parallel photonic reservoir computing based on frequency multiplexing

Wosen Kassa, Evangelia Dimitriadou, Marc Haelterman, Serge Massar, Erwin Bente

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

Photonic reservoir computing uses recent advances in machine learning, and in particular the reservoir computing algorithm, to carry out complex computations optically. Experimental demonstrations with performance comparable to state of the art digital implementations have been reported. However, most experiments so far were based on sequential processing using time-multiplexing. Parallel architectures promise considerable speedup. Recently, a reservoir computing architecture based on frequency parallelism was proposed by our laboratory, and a preliminary demonstration was carried out using optical fibres. In this system the reservoir is linear and the nonlinearity is provided by readout photodiodes. Here, we study in simulation an implementation of this frequency parallel architecture on an InP chip using a generic integration platform. This would dramatically reduce the footprint and cost of the reservoir. The input signal is encoded by modulating the frequency comb produced by a mode locked laser with a repetition rate of 10GHz. The update rate of the input is 2.5GHz. The reservoir, an active cavity with a time delay of 0.4ns, contains a phase modulator which is driven by a 10GHz RF signal, and a semiconductor amplifier to compensate the losses in the cavity. Readout is carried out by measuring the intensity of individual frequency combs and linearly combining them. We performed time domain simulations on a standard channel equalization task. The simulation takes in to account the phase and amplitude noise of the laser source, and the amplifier noise. The power leakage between neighboring channels at the de-multiplexer is also included. To evaluate the system performance, noise is added as a global parameter on the input signal to assess the SNR requirements. Simulation results show that the laser phase noise is far more important that other types of noise, hence the laser source design/operation should be optimized to achieve low phase noise comb.

Original languageEnglish
Title of host publicationNeuro-inspired Photonic Computing
EditorsPeter Bienstman, Marc Sciamanna
PublisherSPIE
Number of pages7
Volume10689
ISBN (Print)9781510619043
DOIs
Publication statusPublished - 1 Jan 2018
EventSPIE Photonics Europe: Neuro-inspired Photonic Computing 2018 - Strasbourg, France
Duration: 22 Apr 201826 Apr 2018
Conference number: 1068903
http://spie.org/conferences-and-exhibitions/photonics-europe?SSO=1

Conference

ConferenceSPIE Photonics Europe: Neuro-inspired Photonic Computing 2018
CountryFrance
CityStrasbourg
Period22/04/1826/04/18
Internet address

Fingerprint

Multiplexing
multiplexing
Photonics
photonics
Parallel architectures
Lasers
Computing
Phase noise
Phase Noise
Demonstrations
Parallel Architectures
Laser
lasers
readout
Cavity
Simulation
simulation
Laser modes
amplifiers
Photodiodes

Cite this

Kassa, W., Dimitriadou, E., Haelterman, M., Massar, S., & Bente, E. (2018). Towards integrated parallel photonic reservoir computing based on frequency multiplexing. In P. Bienstman, & M. Sciamanna (Eds.), Neuro-inspired Photonic Computing (Vol. 10689). [1068903] SPIE. https://doi.org/10.1117/12.2306176
Kassa, Wosen ; Dimitriadou, Evangelia ; Haelterman, Marc ; Massar, Serge ; Bente, Erwin. / Towards integrated parallel photonic reservoir computing based on frequency multiplexing. Neuro-inspired Photonic Computing. editor / Peter Bienstman ; Marc Sciamanna. Vol. 10689 SPIE, 2018.
@inproceedings{a20e76922aeb496691a34eda0e0c8f43,
title = "Towards integrated parallel photonic reservoir computing based on frequency multiplexing",
abstract = "Photonic reservoir computing uses recent advances in machine learning, and in particular the reservoir computing algorithm, to carry out complex computations optically. Experimental demonstrations with performance comparable to state of the art digital implementations have been reported. However, most experiments so far were based on sequential processing using time-multiplexing. Parallel architectures promise considerable speedup. Recently, a reservoir computing architecture based on frequency parallelism was proposed by our laboratory, and a preliminary demonstration was carried out using optical fibres. In this system the reservoir is linear and the nonlinearity is provided by readout photodiodes. Here, we study in simulation an implementation of this frequency parallel architecture on an InP chip using a generic integration platform. This would dramatically reduce the footprint and cost of the reservoir. The input signal is encoded by modulating the frequency comb produced by a mode locked laser with a repetition rate of 10GHz. The update rate of the input is 2.5GHz. The reservoir, an active cavity with a time delay of 0.4ns, contains a phase modulator which is driven by a 10GHz RF signal, and a semiconductor amplifier to compensate the losses in the cavity. Readout is carried out by measuring the intensity of individual frequency combs and linearly combining them. We performed time domain simulations on a standard channel equalization task. The simulation takes in to account the phase and amplitude noise of the laser source, and the amplifier noise. The power leakage between neighboring channels at the de-multiplexer is also included. To evaluate the system performance, noise is added as a global parameter on the input signal to assess the SNR requirements. Simulation results show that the laser phase noise is far more important that other types of noise, hence the laser source design/operation should be optimized to achieve low phase noise comb.",
author = "Wosen Kassa and Evangelia Dimitriadou and Marc Haelterman and Serge Massar and Erwin Bente",
year = "2018",
month = "1",
day = "1",
doi = "10.1117/12.2306176",
language = "English",
isbn = "9781510619043",
volume = "10689",
editor = "Peter Bienstman and Marc Sciamanna",
booktitle = "Neuro-inspired Photonic Computing",
publisher = "SPIE",
address = "United States",

}

Kassa, W, Dimitriadou, E, Haelterman, M, Massar, S & Bente, E 2018, Towards integrated parallel photonic reservoir computing based on frequency multiplexing. in P Bienstman & M Sciamanna (eds), Neuro-inspired Photonic Computing. vol. 10689, 1068903, SPIE, SPIE Photonics Europe: Neuro-inspired Photonic Computing 2018, Strasbourg, France, 22/04/18. https://doi.org/10.1117/12.2306176

Towards integrated parallel photonic reservoir computing based on frequency multiplexing. / Kassa, Wosen; Dimitriadou, Evangelia; Haelterman, Marc; Massar, Serge; Bente, Erwin.

Neuro-inspired Photonic Computing. ed. / Peter Bienstman; Marc Sciamanna. Vol. 10689 SPIE, 2018. 1068903.

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - Towards integrated parallel photonic reservoir computing based on frequency multiplexing

AU - Kassa, Wosen

AU - Dimitriadou, Evangelia

AU - Haelterman, Marc

AU - Massar, Serge

AU - Bente, Erwin

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Photonic reservoir computing uses recent advances in machine learning, and in particular the reservoir computing algorithm, to carry out complex computations optically. Experimental demonstrations with performance comparable to state of the art digital implementations have been reported. However, most experiments so far were based on sequential processing using time-multiplexing. Parallel architectures promise considerable speedup. Recently, a reservoir computing architecture based on frequency parallelism was proposed by our laboratory, and a preliminary demonstration was carried out using optical fibres. In this system the reservoir is linear and the nonlinearity is provided by readout photodiodes. Here, we study in simulation an implementation of this frequency parallel architecture on an InP chip using a generic integration platform. This would dramatically reduce the footprint and cost of the reservoir. The input signal is encoded by modulating the frequency comb produced by a mode locked laser with a repetition rate of 10GHz. The update rate of the input is 2.5GHz. The reservoir, an active cavity with a time delay of 0.4ns, contains a phase modulator which is driven by a 10GHz RF signal, and a semiconductor amplifier to compensate the losses in the cavity. Readout is carried out by measuring the intensity of individual frequency combs and linearly combining them. We performed time domain simulations on a standard channel equalization task. The simulation takes in to account the phase and amplitude noise of the laser source, and the amplifier noise. The power leakage between neighboring channels at the de-multiplexer is also included. To evaluate the system performance, noise is added as a global parameter on the input signal to assess the SNR requirements. Simulation results show that the laser phase noise is far more important that other types of noise, hence the laser source design/operation should be optimized to achieve low phase noise comb.

AB - Photonic reservoir computing uses recent advances in machine learning, and in particular the reservoir computing algorithm, to carry out complex computations optically. Experimental demonstrations with performance comparable to state of the art digital implementations have been reported. However, most experiments so far were based on sequential processing using time-multiplexing. Parallel architectures promise considerable speedup. Recently, a reservoir computing architecture based on frequency parallelism was proposed by our laboratory, and a preliminary demonstration was carried out using optical fibres. In this system the reservoir is linear and the nonlinearity is provided by readout photodiodes. Here, we study in simulation an implementation of this frequency parallel architecture on an InP chip using a generic integration platform. This would dramatically reduce the footprint and cost of the reservoir. The input signal is encoded by modulating the frequency comb produced by a mode locked laser with a repetition rate of 10GHz. The update rate of the input is 2.5GHz. The reservoir, an active cavity with a time delay of 0.4ns, contains a phase modulator which is driven by a 10GHz RF signal, and a semiconductor amplifier to compensate the losses in the cavity. Readout is carried out by measuring the intensity of individual frequency combs and linearly combining them. We performed time domain simulations on a standard channel equalization task. The simulation takes in to account the phase and amplitude noise of the laser source, and the amplifier noise. The power leakage between neighboring channels at the de-multiplexer is also included. To evaluate the system performance, noise is added as a global parameter on the input signal to assess the SNR requirements. Simulation results show that the laser phase noise is far more important that other types of noise, hence the laser source design/operation should be optimized to achieve low phase noise comb.

UR - http://www.scopus.com/inward/record.url?scp=85053437998&partnerID=8YFLogxK

U2 - 10.1117/12.2306176

DO - 10.1117/12.2306176

M3 - Conference contribution

AN - SCOPUS:85053437998

SN - 9781510619043

VL - 10689

BT - Neuro-inspired Photonic Computing

A2 - Bienstman, Peter

A2 - Sciamanna, Marc

PB - SPIE

ER -

Kassa W, Dimitriadou E, Haelterman M, Massar S, Bente E. Towards integrated parallel photonic reservoir computing based on frequency multiplexing. In Bienstman P, Sciamanna M, editors, Neuro-inspired Photonic Computing. Vol. 10689. SPIE. 2018. 1068903 https://doi.org/10.1117/12.2306176