• DocumentCode
    3561337
  • Title

    Empirical Volterra-Series Modeling of Commercial Light-Emitting Diodes

  • Author

    Kamalakis, Thomas ; Walewski, Joachim W. ; Ntogari, Georgia ; Mileounis, Gerasimos

  • Author_Institution
    Dept. of Inf. & Telematics, Harokopio Univ., Athens, Greece
  • Volume
    29
  • Issue
    14
  • fYear
    2011
  • fDate
    7/15/2011 12:00:00 AM
  • Firstpage
    2146
  • Lastpage
    2155
  • Abstract
    Light-emitting diodes (LEDs) constitute a low-cost alternative for optical data transmission of up to ~ 1 Gb/s. What differentiates such applications from, e.g., backhaul optical networks, is the fact that apart from their data throughput, LEDs are generally not as well characterized by the manufacturer as, for example, optical fiber amplifiers. While for simple modulation formats, this lack of knowledge is not a severe impediment; in any other situation, one may face rather complex behaviors of commercial LEDs. In this paper, the main electro-optical characteristics of LEDs are discussed, and it is shown that some popular simple nonlinear models available in the literature are inadequate in describing their dynamics. As a way out of this malady, we present a reverse-engineering approach that is based on Volterra expansions of the electro-optical characteristic function of LEDs, enabling the introduction of a realistic empirical model for commercial devices.
  • Keywords
    Volterra series; electro-optical devices; light emitting diodes; optical communication equipment; semiconductor device models; LED; electro-optical characteristics; empirical Volterra-series modeling; light-emitting diodes; modulation formats; optical data transmission; Current measurement; Frequency measurement; Kernel; Light emitting diodes; Mathematical model; Nonlinear optics; Optical variables measurement; Electro-optic modulation; Volterra series; light-emitting diodes (LEDs); nonlinearities;
  • fLanguage
    English
  • Journal_Title
    Lightwave Technology, Journal of
  • Publisher
    ieee
  • Conference_Location
    5/19/2011 12:00:00 AM
  • ISSN
    0733-8724
  • Type

    jour

  • DOI
    10.1109/JLT.2011.2157082
  • Filename
    5771519