G. Sefler, U. Paudel, T. J. Shaw, D. Monahan, A. Scofield, S. Estella, L. Johansson, G. Valley
{"title":"Laser Speckle in Multimode Waveguides for Random Projections in Compressive Sensing and Reservoir Computing","authors":"G. Sefler, U. Paudel, T. J. Shaw, D. Monahan, A. Scofield, S. Estella, L. Johansson, G. Valley","doi":"10.1109/IPCon.2019.8908273","DOIUrl":null,"url":null,"abstract":"Detection of wideband RF signals has applications in sensing and communications. When the signals of interest are sparse, compressive sensing (CS) provides a sub-Nyquist sampling strategy with potential size, weight, and power savings. The critical element in a CS receiver is the device that produces the wideband CS measurement matrix (MM), a MxN matrix with M << N satisfying certain properties [1]. We have shown that passive optical speckle in multimode waveguides provides excellent MMs for CS. The M rows of the MM are obtained from M photodetectors placed at different locations within the output speckle pattern. A range of algorithms can be used to recover the sparse input signal from the resulting measurement vector. We have experimentally demonstrated two speckle-based CS systems: (1) a real-time system with M = 16 implemented using multimode fiber (MMF) that recovers RF frequency, amplitude, and phase, and (2) a simplified spectrometer system implemented using a multimode planar waveguide on a silicon photonic chip that detects only RF frequency and amplitude.","PeriodicalId":314151,"journal":{"name":"2019 IEEE Photonics Conference (IPC)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 IEEE Photonics Conference (IPC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/IPCon.2019.8908273","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Detection of wideband RF signals has applications in sensing and communications. When the signals of interest are sparse, compressive sensing (CS) provides a sub-Nyquist sampling strategy with potential size, weight, and power savings. The critical element in a CS receiver is the device that produces the wideband CS measurement matrix (MM), a MxN matrix with M << N satisfying certain properties [1]. We have shown that passive optical speckle in multimode waveguides provides excellent MMs for CS. The M rows of the MM are obtained from M photodetectors placed at different locations within the output speckle pattern. A range of algorithms can be used to recover the sparse input signal from the resulting measurement vector. We have experimentally demonstrated two speckle-based CS systems: (1) a real-time system with M = 16 implemented using multimode fiber (MMF) that recovers RF frequency, amplitude, and phase, and (2) a simplified spectrometer system implemented using a multimode planar waveguide on a silicon photonic chip that detects only RF frequency and amplitude.