Computing matrix inversion with optical networks
Computing matrix inversion with optical networks
With this paper we bring about a discussion on the computing potential of complex optical networks and provide experimental demonstration that an optical fiber network can be used as an analog processor to calculate matrix inversion. A 3x3 matrix is inverted as a proof-of-concept demonstration using a fiber network containing three nodes and operating at telecomm wavelength. For an NxN matrix, the overall solving time (including setting time of the matrix elements and calculation time of inversion) scales as O(N2), whereas matrix inversion by most advanced computer algorithms requires ~O(N2.37) computational time. For well-conditioned matrices, the error of the inversion performed optically is found to be around 3%, limited by the accuracy of measurement equipment.
295-304
Wu, Kan
2257a2a5-e70b-461d-8ae5-31b98ef572e5
Soci, Cesare
6c86324e-2968-4e90-9436-4a92a4b26cec
Shum, Perry Ping
5c669130-28f6-4b50-8568-78e05e2ce87d
Zheludev, Nikolay I.
32fb6af7-97e4-4d11-bca6-805745e40cc6
13 January 2014
Wu, Kan
2257a2a5-e70b-461d-8ae5-31b98ef572e5
Soci, Cesare
6c86324e-2968-4e90-9436-4a92a4b26cec
Shum, Perry Ping
5c669130-28f6-4b50-8568-78e05e2ce87d
Zheludev, Nikolay I.
32fb6af7-97e4-4d11-bca6-805745e40cc6
Wu, Kan, Soci, Cesare, Shum, Perry Ping and Zheludev, Nikolay I.
(2014)
Computing matrix inversion with optical networks.
Optics Express, 22 (1), .
(doi:10.1364/OE.22.000295).
Abstract
With this paper we bring about a discussion on the computing potential of complex optical networks and provide experimental demonstration that an optical fiber network can be used as an analog processor to calculate matrix inversion. A 3x3 matrix is inverted as a proof-of-concept demonstration using a fiber network containing three nodes and operating at telecomm wavelength. For an NxN matrix, the overall solving time (including setting time of the matrix elements and calculation time of inversion) scales as O(N2), whereas matrix inversion by most advanced computer algorithms requires ~O(N2.37) computational time. For well-conditioned matrices, the error of the inversion performed optically is found to be around 3%, limited by the accuracy of measurement equipment.
Text
oe-22-1-295.pdf
- Version of Record
Available under License Other.
More information
e-pub ahead of print date: 2 January 2014
Published date: 13 January 2014
Additional Information:
Funded by Singapore Ministry of Education: Singapore Ministry of Education Academic Research Fund Tier 3 (MOE2011-T3-1-005)
Organisations:
Optoelectronics Research Centre
Identifiers
Local EPrints ID: 369144
URI: http://eprints.soton.ac.uk/id/eprint/369144
ISSN: 1094-4087
PURE UUID: 28d04251-e240-4014-bcb0-f3a5cbd967a4
Catalogue record
Date deposited: 18 Sep 2014 13:29
Last modified: 15 Mar 2024 02:44
Export record
Altmetrics
Contributors
Author:
Kan Wu
Author:
Cesare Soci
Author:
Perry Ping Shum
Author:
Nikolay I. Zheludev
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics