Alessandro Lupo, Enrico Picco, Marina Zajnulina, and Serge Massar
{"title":"基于层间全模拟连接频率复用的深层光子储层计算机","authors":"Alessandro Lupo, Enrico Picco, Marina Zajnulina, and Serge Massar","doi":"10.1364/optica.489501","DOIUrl":null,"url":null,"abstract":"Reservoir computers (RCs) are randomized recurrent neural networks well adapted to process time series, performing tasks such as nonlinear distortion compensation or prediction of chaotic dynamics. Deep reservoir computers (deep-RCs), in which the output of one reservoir is used as the input for another one, can lead to improved performance because, as in other deep artificial neural networks, the successive layers represent the data in more and more abstract ways. We present a fiber-based photonic implementation of a two-layer deep-RC based on frequency multiplexing. The two RC layers are encoded in two frequency combs propagating in the same experimental setup. The connection between the layers is fully analog and does not require any digital processing. We find that the deep-RC outperforms a traditional RC by up to two orders of magnitude on two benchmark tasks. This work paves the way towards using fully analog photonic neuromorphic computing for complex processing of time series, while avoiding costly analog-to-digital and digital-to-analog conversions.","PeriodicalId":19515,"journal":{"name":"Optica","volume":"22 3","pages":""},"PeriodicalIF":8.4000,"publicationDate":"2023-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Deep photonic reservoir computer based on frequency multiplexing with fully analog connection between layers\",\"authors\":\"Alessandro Lupo, Enrico Picco, Marina Zajnulina, and Serge Massar\",\"doi\":\"10.1364/optica.489501\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Reservoir computers (RCs) are randomized recurrent neural networks well adapted to process time series, performing tasks such as nonlinear distortion compensation or prediction of chaotic dynamics. Deep reservoir computers (deep-RCs), in which the output of one reservoir is used as the input for another one, can lead to improved performance because, as in other deep artificial neural networks, the successive layers represent the data in more and more abstract ways. We present a fiber-based photonic implementation of a two-layer deep-RC based on frequency multiplexing. The two RC layers are encoded in two frequency combs propagating in the same experimental setup. The connection between the layers is fully analog and does not require any digital processing. We find that the deep-RC outperforms a traditional RC by up to two orders of magnitude on two benchmark tasks. This work paves the way towards using fully analog photonic neuromorphic computing for complex processing of time series, while avoiding costly analog-to-digital and digital-to-analog conversions.\",\"PeriodicalId\":19515,\"journal\":{\"name\":\"Optica\",\"volume\":\"22 3\",\"pages\":\"\"},\"PeriodicalIF\":8.4000,\"publicationDate\":\"2023-11-06\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Optica\",\"FirstCategoryId\":\"101\",\"ListUrlMain\":\"https://doi.org/10.1364/optica.489501\",\"RegionNum\":1,\"RegionCategory\":\"物理与天体物理\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"OPTICS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Optica","FirstCategoryId":"101","ListUrlMain":"https://doi.org/10.1364/optica.489501","RegionNum":1,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"OPTICS","Score":null,"Total":0}
Deep photonic reservoir computer based on frequency multiplexing with fully analog connection between layers
Reservoir computers (RCs) are randomized recurrent neural networks well adapted to process time series, performing tasks such as nonlinear distortion compensation or prediction of chaotic dynamics. Deep reservoir computers (deep-RCs), in which the output of one reservoir is used as the input for another one, can lead to improved performance because, as in other deep artificial neural networks, the successive layers represent the data in more and more abstract ways. We present a fiber-based photonic implementation of a two-layer deep-RC based on frequency multiplexing. The two RC layers are encoded in two frequency combs propagating in the same experimental setup. The connection between the layers is fully analog and does not require any digital processing. We find that the deep-RC outperforms a traditional RC by up to two orders of magnitude on two benchmark tasks. This work paves the way towards using fully analog photonic neuromorphic computing for complex processing of time series, while avoiding costly analog-to-digital and digital-to-analog conversions.
期刊介绍:
Optica is an open access, online-only journal published monthly by Optica Publishing Group. It is dedicated to the rapid dissemination of high-impact peer-reviewed research in the field of optics and photonics. The journal provides a forum for theoretical or experimental, fundamental or applied research to be swiftly accessed by the international community. Optica is abstracted and indexed in Chemical Abstracts Service, Current Contents/Physical, Chemical & Earth Sciences, and Science Citation Index Expanded.