Talk:Repetition code

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

The plots are flawed and the article is poor.[edit]

Please make a thorough revision of the article. The plots are wrong! — Preceding unsigned comment added by 88.11.117.80 (talk) 08:12, 5 November 2012 (UTC)[reply]

I do the following changes:

  • Add together the sections "Repetition coder" and "Repetition decoder" as new section "Example". (The sections are short and there is nothing more to say, so better give just one example which contains the same information)
  • Delete sections "Repetition codes on fading channel" and "Repetition codeson Gaussian channel". (First, I claim the results for Gaussian channel are wrong, since a higher repetition factor shouldn't make the results worse. Second, there is no need for simulation results, maybe one could think of probability of error (after decoding) versus bit error probability for different repetition factors. Third, the reader gets confused because of modulation, fading, theoretical/practical bpsk, SNR, etc. which is too much detail for repetition code.)
  • Revision of section "Code parameters". (Too many parameters and symbols e.g. Hamming distance (), minimum weight () are unnecessarily introduced, since it's such a simple code and this makes it more complicated as it is. Maybe one could move the content of this section to the general description.)
  • Delete parts of section "Applications". (First, they find applications in all environments not only in fading and non-AWGN, but in general they play a minor role. Second, as every error correction code they can be viewed as a method of time-diversity, but space-time-diversity? So either wrong or not necessary since it's a property for error correction codes in general. Third, the analogy to repetition of many cylces of sinusoids is either wrong or that much rocket science, that it's wrong placed here.)
  • Deletion of section "See also". (Link to "majority logic decoding" and "Hamming code" is now in the text, between "Convolutional code" and repetition code is no special connection as well as to "triple modular redundancy".)
  • Revision of other sections.

--BBC89 (talk) 08:24, 5 April 2017 (UTC)[reply]

Misleading example[edit]

Do you think the example is a bit misleading, regarding terminology? Because both the repetition code length and the number of transmitted bits are equal to three.

sissyneck (talk) 10:27, 14 August 2018 (UTC)[reply]