A multi-mode entropy decoder with a generic table partition strategy

Jui Chin Chu*, Liang Fei Su, Yao Chang Yang, Jiun-In  Guo, Ching Lung Su

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

In this paper a low cost and low power multi-mode entropy decoder is proposed. The proposed design is compatible to the entropy decoding for JPEG, MPEG-1/2/4, H.264 and VC-1 video coding standards. It adopts the code-word tables merging and sharing, and integrates the various entropy decoding into a single programmable design. To reduce the required memory space, a generic look-up table partition strategy covering various video coding standards is proposed. Besides, the low power concept of high probability data path with lower capacitance is also taken into account. The proposed multi-mode entropy decoder is implemented using TSMC 0.13μm at the cost of 113,884 gates and 0.54KB SRAM. Its maximum operating frequency achieves 166MHz, which can support entropy decoding on high definition video larger than HD1080@48fps.

Original languageEnglish
Title of host publication2008 International SoC Design Conference, ISOCC 2008
DOIs
StatePublished - 1 Dec 2008
Event2008 International SoC Design Conference, ISOCC 2008 - Busan, Korea, Republic of
Duration: 24 Nov 200825 Nov 2008

Publication series

Name2008 International SoC Design Conference, ISOCC 2008
Volume1

Conference

Conference2008 International SoC Design Conference, ISOCC 2008
CountryKorea, Republic of
CityBusan
Period24/11/0825/11/08

Keywords

  • H.264
  • JPEG
  • MPEG-1/2/4
  • Multiple standard video decoder
  • Variable length decoding (VLD)
  • VC-1

Fingerprint Dive into the research topics of 'A multi-mode entropy decoder with a generic table partition strategy'. Together they form a unique fingerprint.

Cite this