• DocumentCode
    549623
  • Title

    Energy-efficient MIMO detection using unequal error protection for embedded joint decoding system

  • Author

    Yang, Yoon Seok ; Bhagawat, Pankaj ; Choi, Gwan

  • Author_Institution
    Electr. & Comput. Eng., Texas A&M Univ., College Station, TX, USA
  • fYear
    2011
  • fDate
    5-9 June 2011
  • Firstpage
    579
  • Lastpage
    584
  • Abstract
    This paper presents a low-power multiple-input multiple-output (MIMO) and H.264 video joint detector/decoder design that minimizes energy for portable, wireless embedded systems. The design combines an unequal error protection (UEP) scheme with a variable fidelity of MIMO detection, making performance and energy consumption tradeoffs as required by the importance of each coded video frame. Using search space reduction/truncation strategies in MIMO detection along with H.264 data partitioning (DP) method, we determine the decoding configurations that yield minimum energy consumption for pre-specified image qualities at the receiver. The synthesis result of this scalable MIMO detector using a 45nm technology library yields 52%, 57%, and 58% energy reductions at 24dB SNR in Foreman, Akiyo, and Mobile test streams respectively. The tradeoff is negligible 0.3dB degradation in peak signal to noise ratio (PSNR) typically used in an image quality estimation.
  • Keywords
    MIMO communication; decoding; embedded systems; low-power electronics; video coding; H.264 data partitioning method; H.264 video joint detector/decoder design; embedded joint decoding system; energy-efficient MIMO detection; image qualities; low-power multiple-input multiple-output; portable wireless embedded systems; search space reduction-truncation strategies; size 45 nm; unequal error protection; Decoding; Energy consumption; Joints; MIMO; Mobile communication; PSNR; Streaming media; Unequal error protection; low power design; wireless system;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Design Automation Conference (DAC), 2011 48th ACM/EDAC/IEEE
  • Conference_Location
    New York, NY
  • ISSN
    0738-100x
  • Print_ISBN
    978-1-4503-0636-2
  • Type

    conf

  • Filename
    5981979