File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Rotating neurons for all-analog implementation of cyclic reservoir computing

TitleRotating neurons for all-analog implementation of cyclic reservoir computing
Authors
Issue Date2022
Citation
Nature Communications, 2022, v. 13, n. 1, article no. 1549 How to Cite?
AbstractHardware implementation in resource-efficient reservoir computing is of great interest for neuromorphic engineering. Recently, various devices have been explored to implement hardware-based reservoirs. However, most studies were mainly focused on the reservoir layer, whereas an end-to-end reservoir architecture has yet to be developed. Here, we propose a versatile method for implementing cyclic reservoirs using rotating elements integrated with signal-driven dynamic neurons, whose equivalence to standard cyclic reservoir algorithm is mathematically proven. Simulations show that the rotating neuron reservoir achieves record-low errors in a nonlinear system approximation benchmark. Furthermore, a hardware prototype was developed for near-sensor computing, chaotic time-series prediction and handwriting classification. By integrating a memristor array as a fully-connected output layer, the all-analog reservoir computing system achieves 94.0% accuracy, while simulation shows >1000× lower system-level power than prior works. Therefore, our work demonstrates an elegant rotation-based architecture that explores hardware physics as computational resources for high-performance reservoir computing.
Persistent Identifierhttp://hdl.handle.net/10722/334818
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorLiang, Xiangpeng-
dc.contributor.authorZhong, Yanan-
dc.contributor.authorTang, Jianshi-
dc.contributor.authorLiu, Zhengwu-
dc.contributor.authorYao, Peng-
dc.contributor.authorSun, Keyang-
dc.contributor.authorZhang, Qingtian-
dc.contributor.authorGao, Bin-
dc.contributor.authorHeidari, Hadi-
dc.contributor.authorQian, He-
dc.contributor.authorWu, Huaqiang-
dc.date.accessioned2023-10-20T06:50:58Z-
dc.date.available2023-10-20T06:50:58Z-
dc.date.issued2022-
dc.identifier.citationNature Communications, 2022, v. 13, n. 1, article no. 1549-
dc.identifier.urihttp://hdl.handle.net/10722/334818-
dc.description.abstractHardware implementation in resource-efficient reservoir computing is of great interest for neuromorphic engineering. Recently, various devices have been explored to implement hardware-based reservoirs. However, most studies were mainly focused on the reservoir layer, whereas an end-to-end reservoir architecture has yet to be developed. Here, we propose a versatile method for implementing cyclic reservoirs using rotating elements integrated with signal-driven dynamic neurons, whose equivalence to standard cyclic reservoir algorithm is mathematically proven. Simulations show that the rotating neuron reservoir achieves record-low errors in a nonlinear system approximation benchmark. Furthermore, a hardware prototype was developed for near-sensor computing, chaotic time-series prediction and handwriting classification. By integrating a memristor array as a fully-connected output layer, the all-analog reservoir computing system achieves 94.0% accuracy, while simulation shows >1000× lower system-level power than prior works. Therefore, our work demonstrates an elegant rotation-based architecture that explores hardware physics as computational resources for high-performance reservoir computing.-
dc.languageeng-
dc.relation.ispartofNature Communications-
dc.titleRotating neurons for all-analog implementation of cyclic reservoir computing-
dc.typeArticle-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1038/s41467-022-29260-1-
dc.identifier.pmid35322037-
dc.identifier.scopuseid_2-s2.0-85126889090-
dc.identifier.volume13-
dc.identifier.issue1-
dc.identifier.spagearticle no. 1549-
dc.identifier.epagearticle no. 1549-
dc.identifier.eissn2041-1723-
dc.identifier.isiWOS:000772575300022-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats