Demultiplexing through a multimode fiber using chip-scale diffractive neural networks

Demultiplexing through a multimode fiber using chip-scale diffractive neural networks | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Demultiplexing through a multimode fiber using chip-scale diffractive neural networks Qian Zhang, Haoyi Yu, Jie Zhang, Yuedi Zhang, Chao Meng, Jiali Sun, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8309375/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The explosive growth of global data traffic is pushing conventional single-mode fiber systems toward their fundamental capacity limits. Multimode fibers (MMFs) using space-division multiplexing (SDM) are promising for improved transmission capacity, connection flexibility, and security of data. However, the complex transmission characteristics of MMFs significantly hinder precise mode demultiplexing. Conventional approaches, such as holographic measurements, phase recovery algorithms, photonic lanterns, and multi-plane light conversion, are severely limited by instability, ambiguity, flexibility, or the complexity of the system. In this paper, we demonstrate for the first time a purely optical, chip-scale AI solution for high-mode isolation, speed-of-light demultiplexing of MMF modes using a three-dimensional diffractive neural network (DNN). The DNN is trained with synthetic modal data and fabricated using two-photon nanolithography. It features a compact size of 120μm ×120μm ×80μm and a diffractive structure size of 1μm for the neurons at the hidden layers of the network. The proposed DNN-based demultiplexer achieves a relative modal amplitude error of 6.93% in simulation and 17.96% in experiment. The AI approach of DNN allows for flexible design and overcomes the size and performance limitations of digital-optical demultiplexers. This work paves the way for compact, low-latency optical processors for high-performance demultiplexers and enables scalable, chip-integrated solutions for next-generation fiber optic networks. Physical sciences/Physics/Electronics, photonics and device physics Physical sciences/Optics and photonics/Applied optics/Fibre optics and optical communications Full Text Additional Declarations There is NO Competing Interest. Supplementary Files DNNoutputrealtime.mp4 Real-time output of DNN-based DEMUX DNNsupplementsubmission.pdf Supplementary Material Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8309375","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":557600054,"identity":"efb0b218-9c1b-4d45-abc0-ae099e555cc7","order_by":0,"name":"Qian Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABDUlEQVRIiWNgGAWjYDACZjB5AMQ6AOMawMUJaGFLbCBOCwNcC48hcVr4jjM/e8Dw5468Of+a7495GKzz+fkPb3xcwGAth0uL5GE2cwMGnmeGO2e83djMw5BuOXNGWrHxDIZ0Y1xaDA4zmEkwSBxm3HDj7MZm3n+HDQxu8JhJ8zAcBvkMhxb2bxJA0n7DjTMPgbYcNrA/f8b8N5BRj1sLD9CWhMOJG873MIK1GDDkmDEDGQm4/cJTJpFw4HDyhhtshjPnMKQbSNxIK5bmMUg3xGUL3/nj2yQ+/Dlsu+H84Qcf3jBYG/D3H974mafCWh6XLeAYAbtBAsUlBjg1QLSAAf8B3IpGwSgYBaNgZAMAQChXBQPvb/8AAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0002-1951-2722","institution":"Laboratory of Measurement and Sensor System Technique, TU Dresden","correspondingAuthor":true,"prefix":"","firstName":"Qian","middleName":"","lastName":"Zhang","suffix":""},{"id":557600055,"identity":"9b6e1c2f-6b32-424a-af62-b35670ce6feb","order_by":1,"name":"Haoyi Yu","email":"","orcid":"https://orcid.org/0000-0002-9480-8443","institution":"University of Shanghai for Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Haoyi","middleName":"","lastName":"Yu","suffix":""},{"id":557600056,"identity":"61914cd8-ede9-48c9-af80-0cfced1e2f6a","order_by":2,"name":"Jie Zhang","email":"","orcid":"","institution":"Laboratory of Measurement and Sensor System Technique, TU Dresden","correspondingAuthor":false,"prefix":"","firstName":"Jie","middleName":"","lastName":"Zhang","suffix":""},{"id":557600057,"identity":"74788614-7cbf-40fb-b341-ba0324c72690","order_by":3,"name":"Yuedi Zhang","email":"","orcid":"","institution":"Laboratory of Measurement and Sensor System Technique, TU Dresden","correspondingAuthor":false,"prefix":"","firstName":"Yuedi","middleName":"","lastName":"Zhang","suffix":""},{"id":557600058,"identity":"f9490194-dbc2-4655-9678-058adcfde7d1","order_by":4,"name":"Chao Meng","email":"","orcid":"","institution":"University of Shanghai for Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Chao","middleName":"","lastName":"Meng","suffix":""},{"id":557600059,"identity":"1e3e569c-3687-4777-9bce-3c65c7c42e3f","order_by":5,"name":"Jiali Sun","email":"","orcid":"","institution":"Laboratory of Measurement and Sensor System Technique, TU Dresden","correspondingAuthor":false,"prefix":"","firstName":"Jiali","middleName":"","lastName":"Sun","suffix":""},{"id":557600060,"identity":"19a290c8-9e2e-4cdb-9fb7-a1b4e55f9d4a","order_by":6,"name":"Yu Miao","email":"","orcid":"","institution":"Laboratory of Measurement and Sensor System Technique, TU Dresden","correspondingAuthor":false,"prefix":"","firstName":"Yu","middleName":"","lastName":"Miao","suffix":""},{"id":557600061,"identity":"7e7b0fb9-f1a4-4b52-a2e9-f419d3fa8f76","order_by":7,"name":"Qiming Zhang","email":"","orcid":"https://orcid.org/0000-0003-0067-1210","institution":"USST","correspondingAuthor":false,"prefix":"","firstName":"Qiming","middleName":"","lastName":"Zhang","suffix":""},{"id":557600062,"identity":"dd471807-8ff1-4664-a3f1-5e22f800397c","order_by":8,"name":"Min Gu","email":"","orcid":"https://orcid.org/0000-0003-4078-253X","institution":"University of Shanghai for Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Min","middleName":"","lastName":"Gu","suffix":""},{"id":557600063,"identity":"6e9515bc-e769-494e-a8f1-d09a48b5040d","order_by":9,"name":"Juergen Czarske","email":"","orcid":"https://orcid.org/0000-0001-7280-0523","institution":"TU Dresden","correspondingAuthor":false,"prefix":"","firstName":"Juergen","middleName":"","lastName":"Czarske","suffix":""}],"badges":[],"createdAt":"2025-12-08 15:40:57","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8309375/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8309375/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97840428,"identity":"3f3fe303-d25d-4b60-9b23-9f52d266d1a6","added_by":"auto","created_at":"2025-12-10 03:39:32","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8519782,"visible":true,"origin":"","legend":"","description":"","filename":"DNNsubmission.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8309375/v1/bb6b8f8db1a7357fa6c5d990.pdf"},{"id":97840424,"identity":"3f790828-9b30-459e-b370-d607fdb5f48f","added_by":"auto","created_at":"2025-12-10 03:39:30","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11018,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25099346.json","url":"https://assets-eu.researchsquare.com/files/rs-8309375/v1/e7979edc6d92b76e41a9139c.json"},{"id":97840429,"identity":"1ae6572b-3811-4281-a135-33d81534f678","added_by":"auto","created_at":"2025-12-10 03:39:32","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1309253,"visible":true,"origin":"","legend":"","description":"","filename":"DNNsupplementsubmission.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8309375/v1/e5a5eca078770495265e5895.pdf"},{"id":105564966,"identity":"58a6ccad-e8a9-4719-af36-50d7e3482b26","added_by":"auto","created_at":"2026-03-27 12:51:28","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3069266,"visible":true,"origin":"","legend":"Article File","description":"","filename":"DNNsubmission.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8309375/v1_covered_f5bf6bdb-42f8-44e6-9165-e5462097c59c.pdf"},{"id":97840427,"identity":"08198e1e-4a35-438d-8326-4e9de691a31f","added_by":"auto","created_at":"2025-12-10 03:39:32","extension":"mp4","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":672111,"visible":true,"origin":"","legend":"Real-time output of DNN-based DEMUX","description":"","filename":"DNNoutputrealtime.mp4","url":"https://assets-eu.researchsquare.com/files/rs-8309375/v1/75c64ec7c134ffcc4934a292.mp4"},{"id":97840423,"identity":"de801111-2244-4d59-b30a-d3d89e15902e","added_by":"auto","created_at":"2025-12-10 03:39:30","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":1309253,"visible":true,"origin":"","legend":"Supplementary Material","description":"","filename":"DNNsupplementsubmission.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8309375/v1/4b903d817ebeba3eb75f7bbc.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Demultiplexing through a multimode fiber using chip-scale diffractive neural networks","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8309375/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8309375/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"The explosive growth of global data traffic is pushing conventional single-mode fiber systems toward their fundamental capacity limits. Multimode fibers (MMFs) using space-division multiplexing (SDM) are promising for improved transmission capacity, connection flexibility, and security of data. However, the complex transmission characteristics of MMFs significantly hinder precise mode demultiplexing. Conventional approaches, such as holographic measurements, phase recovery algorithms, photonic lanterns, and multi-plane light conversion, are severely limited by instability, ambiguity, flexibility, or the complexity of the system. In this paper, we demonstrate for the first time a purely optical, chip-scale AI solution for high-mode isolation, speed-of-light demultiplexing of MMF modes using a three-dimensional diffractive neural network (DNN). The DNN is trained with synthetic modal data and fabricated using two-photon nanolithography. It features a compact size of 120μm ×120μm ×80μm and a diffractive structure size of 1μm\u003c2\u003e for the neurons at the hidden layers of the network. The proposed DNN-based demultiplexer achieves a relative modal amplitude error of 6.93% in simulation and 17.96% in experiment. The AI approach of DNN allows for flexible design and overcomes the size and performance limitations of digital-optical demultiplexers. This work paves the way for compact, low-latency optical processors for high-performance demultiplexers and enables scalable, chip-integrated solutions for next-generation fiber optic networks.","manuscriptTitle":"Demultiplexing through a multimode fiber using chip-scale diffractive neural networks","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-10 03:38:28","doi":"10.21203/rs.3.rs-8309375/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ee93db42-a6b4-494c-9cc2-54263e8caecc","owner":[],"postedDate":"December 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":59346789,"name":"Physical sciences/Physics/Electronics, photonics and device physics"},{"id":59346790,"name":"Physical sciences/Optics and photonics/Applied optics/Fibre optics and optical communications"}],"tags":[],"updatedAt":"2026-03-24T16:50:15+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-10 03:38:28","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8309375","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8309375","identity":"rs-8309375","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}