• DocumentCode
    3813
  • Title

    Digital Preemphasis in Optical Communication Systems: On the DAC Requirements for Terabit Transmission Applications

  • Author

    Rafique, Danish ; Napoli, Antonio ; Calabro, Stefano ; Spinnler, Bernhard

  • Author_Institution
    , Coriant R&D GmbH, Munich, Germany
  • Volume
    32
  • Issue
    19
  • fYear
    2014
  • fDate
    Oct.1, 1 2014
  • Firstpage
    3247
  • Lastpage
    3256
  • Abstract
    Next-generation coherent optical systems are geared to employ high-speed digital-to-analog converters (DAC), allowing for digital preprocessing of the signal and flexible optical transport networks. However, one of the major obstacles in such architectures is the limited resolution (less than 5.5 effective bits) and –3 dB bandwidth of commercial DACs, typically limited to half of the currently commercial baud rates, and even relatively reduced in case of higher baud rate transponders (400 Gb/s and 1 Tb/s). In this paper, we propose a simple digital preemphasis (DPE) algorithm to compensate for DAC-induced signal distortions, and exhaustively investigate the impact of DAC specifications on system performance, both with and without DPE. As an outcome, performance improvements are established across various DAC hardware requirements (effective number of bits and bandwidth) and channel baud rates, for m-state quadrature amplitude modulation (QAM) formats. In particular, we show that lower order modulation formats are least affected by DAC limitations, however, they benefit the most from DPE in extremely challenging hardware conditions. On the contrary, higher order formats are severely limited by DAC distortions, and moderately benefit from DPE across a wide range of DAC specifications. Moreover, effective number of bit requirements are established for m-state QAM, assuming low and high baud rate transmission regimes. Finally, we discuss the application scenarios for the proposed DPE in next-generation terabit transmission systems, and establish maximum transportable baud rates, which are shown to be used toward increasing channel baud rates to reduce terabit subcarrier count or toward increasing forward error correction (FEC) overheads to reduce the pre-FEC bit error rate threshold. Maximum baud rates after DPE are summarized here for polarization multiplexed BPSK, QPSK, 8QAM, and 16QAM, assuming two DACs: Current commer- ial DACs (5.5 effective bits, 16 GHz bandwidth) 57, 54, 51, and 48 Gbaud, respectively. Next-generation DACs (7 effective bits, 22 GHz bandwidth): 62, 61, 60, and 58 Gbaud, respectively.
  • Keywords
    Bandwidth; Noise; Q-factor; Quadrature amplitude modulation; Receivers; Transfer functions; Coherent detection; Nyquist; digital signal processing; digital-to-analog converter; pre-emphasis;
  • fLanguage
    English
  • Journal_Title
    Lightwave Technology, Journal of
  • Publisher
    ieee
  • ISSN
    0733-8724
  • Type

    jour

  • DOI
    10.1109/JLT.2014.2343957
  • Filename
    6868202