Re: I know this won't work, I just want to know -why-.




"Claudio Grondi" <claudio.grondi@xxxxxxxxxx> wrote in message
news:don2t4$5b7$1@xxxxxxxxxxxxxxxxxxxxxxxxxxxx
> dalek_no2@xxxxxxxxxxx wrote:
<snip>
>
> Anyone here who has seen files transmitted over the network which got
> damaged due to the way network transmits them and not due to detected
> problems during transmission or failures on storage media?
>
I suspect so.

for example, downloading lots of files over the internet (say, hundreds of
MB or GB worth), lots of errors can pop up (I have seen this happen
sometimes).

there is always some probability that some error will get through
undetected, based mostly on the probability of the error occuring
(mechanical properties, eg, crosstalk, interference, ...) and of the failure
in detecting the error (limitations in the checksum algos, ...).

combined, it can be presumed that with enough data some level of errors are
likely to get through regardless...


as far as error correction, I am not entirely sure how this works reliably,
eg, without "notable" increases in redundency (say, a 3x increase, and
assuming that, eg, 2 out of 3 give the correct answer).

simpler teqniques could be possible, but less reliable, eg, a crc and parity
bits. the parity bits are used to try to locate the damaged byte(s) and the
crc is used to try to "guess" the correct value (likely based on some kind
of hueristic, eg, symbol probabilities and minimal-bit disruptions). in
simple cases (eg: a plain checksum and a single error) one could likely
attempt to calculate the value directly (subtracting the checksum for the
rest of the message from the checksum for the whole message).

then again, to know more I would probably have to actually look into this
kind of thing...

> Claudio


.



Relevant Pages

  • Re: When a Single Photon Hits a Half Silvered Mirror?
    ... It either reflects or it transmits, with a 50-50% probability. ... Neither of these is consistent with experiment. ... That would not conserve momentum. ...
    (sci.physics.relativity)
  • Re: Help: Coding theory and Probability
    ... >> where p is the probability that a digit sent is the digit ... >> I don't really have much of a probability background, ... > I strongly suspect that there may be insufficient information here. ... sender never transmits 111, ...
    (sci.math)
  • [patch 02/32] softmac: remove netif_tx_disable when scanning
    ... In the scan section of ieee80211softmac, network transmits are disabled. ... When SoftMAC re-enables transmits, it may override the wishes of a driver ... that may have very good reasons for disabling transmits. ...
    (Linux-Kernel)
  • [patch 01/24] softmac: remove netif_tx_disable when scanning
    ... In the scan section of ieee80211softmac, network transmits are disabled. ... When SoftMAC re-enables transmits, it may override the wishes of a driver ... that may have very good reasons for disabling transmits. ...
    (Linux-Kernel)