Learning to demultiplex the superposition of spatial modes in multimode fiber using physics-informed deep learning

Research output: Contribution to book/Conference proceedings/Anthology/ReportConference contributionContributedpeer-review

Abstract

We have proposed a physics-informed deep neural network for multimode fiber communication networks. It is able to demultiplex the 55 fiber modes in real-time and without the need for reference light.

Details

Original languageEnglish
Title of host publicationFrontiers in Optics
PublisherOptical Society of America
ISBN (electronic)9781957171296
Publication statusPublished - 2023
Peer-reviewedYes

Conference

Title2023 Frontiers in Optics + Laser Science
Abbreviated titleFiO + LS 2023
Duration9 - 12 October 2023
LocationGreater Tacoma Center
CityTacoma
CountryUnited States of America