Analysis of subpulse generation from delayed-interference signal-wavelength converter for wide carrier recovery rate range

J. Sakaguchi, Mads Lønstrup Nielsen, T. Ohira, R. Suzuki, Y. Ueno

    Research output: Contribution to journalJournal articleResearchpeer-review

    Abstract

    The generation of subpulses in delayed-interference signal-wavelength converters (DISCS) had been suspected as a Curse of waveform degradation in their application in ultrafast (similar to 160GHz) optical time division multiplexing (OTDM) communication systems. We reported its first observation in our previous letter. In this paper, after a detailed explanation of our model. experimental method and improved results with respect to those ill the previous letter, we discussed the applicability of our model in a wide range of Carrier recovery rates ill the semiconductor optical amplifier (SOA). Subpulse waveforms were systematically measured for several carrier recovery rates (carrier lifetimes tau(eff) = 11-280 ps), and an increase in subpulse intensify with recovery rate was observed. The calculated intensities showed good agreement with measured data for most recovery rates, which supports our model. Using this model, we numerically showed the trade-off between subpulses and pattern-induced intensity noise, in the same recovery rate range. This trade-off suggests that part of the DISC structure needs improvement.
    Original languageEnglish
    JournalJapanese Journal of Applied Physics
    Volume47
    Issue number9
    Pages (from-to)7182-7189
    ISSN0021-4922
    DOIs
    Publication statusPublished - 2008

    Keywords

    • pattern-dependent noise
    • semiconductor optical amplifier (SOA)
    • all-optical signal processing
    • delayed-interference signal-wavelength converter (DISC)
    • subpulse
    • nonlinear phase shift
    • wavelength conversion

    Fingerprint

    Dive into the research topics of 'Analysis of subpulse generation from delayed-interference signal-wavelength converter for wide carrier recovery rate range'. Together they form a unique fingerprint.

    Cite this