channel coding and error correction, history.
Because channel coding is somewhat difficult to understand
many try to get going thinking of it as an error correction.
To distuinguish these two types of error reducing mechanisms
terms like
- forward error correction (including channel coding)
- backwards <g> error correction, to ask for retransmission
Forward because the information needed to correct errors
goes "forward" with the actual data.
Backwards because it includes the use of a "backward
channel" carrying that request for retransmission.
Both were originally used with regular streams of bits,
not actual channel, still undetected signals.
The most popular system now is probably the coding on
audio and data CDs, designed to cope with missing bits,
scratches,etc. Those errors can, to a certain degree,
be fixed, calculated from extra "forward error correction"
data on the disk.
Note also the difference between an audio CD having
some bits missing here and there, inspite of the
forward, included, embedded error correction, nobody
normally notices that little extra whatever or missing
whatever.
Compared to the same thing for a data CD, the only
solution is to request a new CD (when there are some
bits missing in the install.exe program or even worse, in
some data file which isn't checked for corruption)
Ilmarinen.
However, it is understandable that many most, at this
point in history, have a problem understanding what
"channel coding" is so that some might try to use
"error correction" instead, but then the whole meaning
of both as different mechanism is lost.
To coin a new word something like "error reducer" might
be better for a channel coder, contain more
linguistic information on what "it" is. |
