6
$\begingroup$

I would think that this would be the case. In any one transmission there will be a systematic error due to the fundamentals of the transmitter / receiver. Is there also a transient or stochastic error from any given transmission that could be effectively eliminated by multiple transmissions of the same data?

Any sources on this would be appreciated.

A somewhat related question: If an orbiter is receiving data from a probe or lander, would it use the primary communication antenna for that, or would it use a different one to keep the primary pointed at Earth?

$\endgroup$

2 Answers 2

7
$\begingroup$

Signals are not repeated, but instead coded in a special way that allows to reconstruct the original data on the receiver side in presence of noise/errors. It is called forward error correction.

FEC schemes are more efficient than just blindly transmitting the same data twice (though they of course increase the total amount of data that has to be transmitted).

$\endgroup$
2
  • 1
    $\begingroup$ Most probes also retain the data in onboard storage until safe receipt on Earth is confirmed. $\endgroup$ Commented Feb 18, 2019 at 8:07
  • 2
    $\begingroup$ Are not usually repeated. I am sure there are occasional missed messages that have to be repeated, due to unusual noise near Jupiter or with the Sun near the line-of-sight. And of course sometimes a spacecraft stops responding and so the message is repeated many many times! See Roberts (SOHO) from here $\endgroup$
    – uhoh
    Commented Feb 18, 2019 at 13:39
4
$\begingroup$

Just a single repetition would not help. If there is a single error within those two data blocks, you do not know where the error is and have no chance to correct the error.

If the data is transmitted three times and you get two identical copies, you may assume those identical copies are correct and one or more errors are in the third block.

A better approach is the use of a code which allows the correction of any single error and the detection of two errors. Variants are possible that allow the correction of several errors and the detection of some more errors. Those codes need less additional bits than the simple dual or triple repetition. Chances of undectable multiple errors could be held very, very low.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.