Noise Margin (an introduction/the literal meaning)
A CANDLE LIT IN JUPITER
Wikipedia didn''t say much about noise margin, and the technical explanations found elsewhere were not very clear for me.
So I consulted an "expert" and tried to understand...
This is the transcription of what was said. It's written in my own words, but aproved by the person who explained it to me (I hope I translated it well). I think it could be useful for all. from it, I got a possible funny title, but I'm still "cooking" my idea for the museum interventions.
Please expect more coming soon, and let me know if something isn't clear.
wikipedia: noise margin is the amount by which a signal exceeds the minimum amount for proper operation.
This means the noise margin is the margin of signal that you have above the level of the noise. Basically if your signal is too close to the noise level, you can not understand it, this is, you can not decode the information contained in the signal (the signal is simply coded information). For example, in a bar there is background noise, this means that many signals (voice or radio signals) coexist in the same space. If you're only interested in one of them, the other become noise to you, because they seem to prevent you from understanding the signal that you want. For example, “A” is in a bar talking to “B”, but there’s too much noise, and they can’t understand each other speaking in a normal volume. “A” can then only do one of two things:
1) “A” can raise her voice level, this means increasing a coefficient called the Signal / Noise ratio (the signal level divided by the noise). The higher this ratio the easier will be for B to understand the signal because the noise will be (comparatively) less annoying.
2) “A” can repeat what she just said. The noise is usually random, and that means that from moment to the next it changes. Because the signal remains unchanged when it is repeated, when joining two signals mixed with noise, the noise is canceled and the signal remains. This is the basic technique of filtering, which consists of adding the signals and using the resulting average. The more you repeat the signal the smaller the average noise becomes.
Going back to what the "noise margin” is, the noise margin is simply a value that tells you how high your signal is over the noise, and thus it’s very similar to the signal / noise ratio. Thus, the higher the signal over noise (in volume, or more technically, in signal power), the fewer problems you'll have to decode. However, nowadays there is great competition for signal transmission. It is as if in a bar everyone wanted to talk at the same time. That is called the frequency spectrum, which is the "space" available for outputting signals. One of the rules for sharing the space is not to "shout" too much: not to use more signal power than strictly necessary. That means that under certain conditions we get the signal with low power, too close to the noise, and that is why digital filtering is used, to eliminate the noise and obtain only the signal. Another limitation to the signal strength may be physical or simply regarding to economic matters: for example a satellite can not deliver high power because it is too far from the earth and it is hugely expensive to mount a big signal amplifier on a satellite. To give you an idea, the GPS signal arrives to the earth so attenuated that it’s like a candle lit in Jupiter!
