Improving the distance properties of turbo codes using a third component code: 3D turbo codes

  • Authors:

  • Claude Berrou;Alexandre Graell i Amat;Youssouf Ould-Cheikh-Mouhamedou;Yannick Saouter

  • Affiliations:

  • Institut TELECOM, TELECOM Bretagne, UMR CNRS Lab-STICC, Université Européenne de Bretagne, Brest Cedex 3, France;Institut TELECOM, TELECOM Bretagne, UMR CNRS Lab-STICC, Université Européenne de Bretagne, Brest Cedex 3, France;Prince Sultan Advanced Technologies Research Institute, King Saud University, Riyadh, Saudi Arabia;Institut TELECOM, TELECOM Bretagne, UMR CNRS Lab-STICC, Université Européenne de Bretagne, Brest Cedex 3, France

  • Venue:
  • IEEE Transactions on Communications
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Thanks to the probabilistic message passing performed between its component decoders, a turbo decoder is able to provide strong error correction close to the theoretical limit. However, the minimum Hamming distance (dmin) of a turbo code may not be sufficiently large to ensure large asymptotic gains at very low error rates (the so-called flattening effect). Increasing the dmin of a turbo code may involve using component encoders with a large number of states, devising more sophisticated internal permutations, or increasing the number of component encoders. This paper addresses the latter option and proposes a modified turbo code in which a fraction of the parity bits are encoded by a rate-1, third encoder. The result is a noticeably increased dmin, which improves turbo decoder performance at low error rates. Performance comparisons with turbo codes and serially concatenated convolutional codes are given.