• DocumentCode
    20892
  • Title

    Designing Optimal Multiresolution Quantizers with Error Detecting Codes

  • Author

    Ho, Jason ; En-Hui Yang

  • Author_Institution
    Dept. of Electr. & Comput. Eng., Univ. of Waterloo, Waterloo, ON, Canada
  • Volume
    12
  • Issue
    7
  • fYear
    2013
  • fDate
    Jul-13
  • Firstpage
    3588
  • Lastpage
    3599
  • Abstract
    This paper investigates the design of optimal multiresolution vector quantizers for broadcast channels with cyclic redundancy checks (CRC). Given a CRC-coded broadcast system with multiresolution vector quantization (MRVQ), a closed-form formula for the weighted end-to-end distortion (EED) is first derived under random index assignment. The closed-form expression is then further utilized to identify necessary optimality conditions to minimize the EED, from which an iterative algorithm is proposed for quantization design. Experiments conducted under both the point-to-point and broadcast channels demonstrate that for a wide range of channel error probability, inclusion of CRC significantly reduces the EED without sacrificing bandwidth. Further analyses are conducted to determine the best tradeoff between bits allocated for source quantization and CRC error detection.
  • Keywords
    broadcast channels; channel coding; cyclic redundancy check codes; error correction codes; iterative decoding; source coding; CRC coded broadcast system; CRC error detection; EED; MRVQ; broadcast channels; channel error probability; closed form formula; closed-form expression; cyclic redundancy checks; designing optimal multiresolution quantizers; error detecting codes; iterative algorithm; multiresolution vector quantization; optimal multiresolution vector quantizer design; point-to-point channels; random index assignment; source quantization; weighted end-to-end distortion; Joint source channel coding; cyclic redundancy check; multiresolution noisy channel quantization;
  • fLanguage
    English
  • Journal_Title
    Wireless Communications, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    1536-1276
  • Type

    jour

  • DOI
    10.1109/TWC.2013.062413.122017
  • Filename
    6552843