| M. Breiling, L. Hanzo, "The Super-Trellis Structure of Turbo Codes". IEEE Transactions on Information Theory, pp. 2212-2228, No. 6, Sept. 2000. |
....For all purposes, it is absolutely mandatory that the interleavers are di#erent. The length of the interleavers, denoted by K, determines the number of information bits encoded into one turbo code word. As maximum likelihood decoding of turbo codes is in general computationally prohibitive [3], we investigate turbo codes together with an iterative decoding algorithm performing (suboptimal) decoding with reasonable complexity. Instead of taking the code constraints of all constituent codes simultaneously into account, each constituent code is decoded separately and passes its output to ....
M. Breiling, L. Hanzo, "The Super-Trellis Structure of Turbo Codes". IEEE Transactions on Information Theory, pp. 2212-2228, No. 6, Sept. 2000.
....is absolutely mandatory that the interleavers within a coding scheme are di#erent. The length of the interleavers, denoted by K, determines the number of information bits encoded into one turbo code word. As maximum likelihood decoding of turbo codes is in general computationally prohibitive [2], we investigate turbo codes together with an iterative decoding algorithm performing (suboptimal) decoding with reasonable complexity. Instead of taking the code constraints of all constituent codes simultaneously into account, each constituent code is decoded separately and passes its output ....
M. Breiling, L. Hanzo, "The Super-Trellis Structure of Turbo Codes". IEEE Transactions on Information Theory, pp. 2212-2228, No. 6, Sept. 2000.
....decoder will have relatively little effect, and the iterative decoding provides good results. Another justification for using the iterative arrangement described above is how well it has been found to work. In the limited experiments that have been carried out with optimal decoding of turbo codes [21] [23] it has been found that optimal decoding performs only a fraction of a decibel (around 0.35 0.5 dB) better than iterative decoding with the MAP algorithm. Furthermore, various turbo coding schemes have been found [23] 24] that approach the Shannonian limit, which gives the best ....
....of turbo codes using BPSK modulation over Rayleigh channels, in the next section we offer our conclusions. XI. CONCLUSION In this article, we have described the techniques used for the decoding of turbo codes. Although it is possible to optimally decode turbo codes in a single noniterative step [21], 22] for complexity reasons a nonoptimum iterative decoder is almost always preferred. Such an iterative decoder employs two component soft in soft out decoders, and we have described the MAP, Log MAP, Max Log MAP and SOVA algorithms, which can all be used as component decoders. The MAP ....
M. Breiling and L. Hanzo, The super-trellis structure of turbo codes, in IEEE Trans. Inform. Theory, Sept. 2000. to appear.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC