How does a single fiber optic cable transmit millions of signals

[jaydnul]

That's how the internet works across oceans, correct? I understand how a fiber optic cable transmits information, but how does it transmit everybody's information at once in just one dollar coin sized cable? One after the other? I find that hard to believe given how fast the internet is, even at the speed of light...

 

[MostlyHarmless]

I was unable to find an exact figure, or anything that directly states this, but I found from a little reading, that there are numerous cables connecting various locations. So it's not just one cable carrying everyone's information to and from other continents. It is several, otherwise, any cable damage(which aren't uncommon from what I've read), be it natural, accidental, or even purposeful, could severely cripple international communications.

The wiki article for "Submarine communications cable" was quite interesting. http://en.wikipedia.org/wiki/Submarine_communications_cable There's even talk of espionage during the cold war! Fascinating stuff!

 

[Drakkith]

One fiber optic cable can handle multiple signals at once. This is called 'Multiplexing'. There are several different ways to multiplex data, but the main one to understand for fiber optics is Wavelength-Division Multiplexing. The short story is that you can send multiple signals by sending them using different wavelengths of light for each signal. See the following link for more info.

http://en.wikipedia.org/wiki/Wavelength-division_multiplexing

 

[CWatters]

It is several, otherwise, any cable damage(which aren't uncommon from what I've read), be it natural, accidental, or even purposeful, could severely cripple international communications.

Even then the odd cable failure causes problems..

http://www.renesys.com/blog/2013/03/intrigue-surrounds-smw4-cut.shtml

This month's submarine cable outages have profoundly degraded connectivity to the Middle East, Asia and Africa. In particular, last week's EASSy and SEACOM outages wiped out connectivity in parts of East Africa from Djibouti to South Africa. As if that weren't bad enough, the biggest submarine cable connecting Europe and Asia, SeaMeWe-4, suffered a failure at 6:20 UTC, 27 March.

Later in the day, the Egyptian Naval forces claimed that they had caught divers sabotaging a submarine cable off the coast of Alexandria, Egypt. In any event, SMW4 clearly suffered a major failure, and as we will see below, caused widespread disruption of Internet services from Egypt to Pakistan.

 

[Crazymechanic]

Also a single cable under it's isolation has multiple little fiber wires each for it's own optical signal.
And because they are fiber optical they can be very tiny compared to a copper wire and so they fit more into a single cable of the same size as previous copper ones.

 

[sophiecentaur]

The smart thing about digital signalling is that, once a signal (voice, data or picture) has been coded into digits, it can be chopped up into short sequences and sent in many different 'packets'. A high bandwidth optical system can handle billions of pulses per second. Your telephone conversation will only need a few thousand pulses per second so you can fit a lot of conversations, pictures or documents into just the one channel by sending all these packets sequentially down the line. The terminal equipment needs to be very clever and to know who each packet is intended for. It is also possible to send more than one wavelength of light down an optical fibre and this further increases the potential capacity of a single fibre.

In the old days, when signals were all analogue, it was still possible to send many different phone conversations down a line by modulating 'carrier waves' with groups of narrow band audio signals (Phone calls are low bandwidth and low quality). There were also methods of switching rapidly between a number of audio signals at the 'send' end and then using a synchronised switch at the other end to sample the emerging signal so that only the wanted signal could be selected. It's always a matter of total available bandwidth being used as efficiently as possible. Digital techniques give you incredibly more effective capacity than the old analogue techniques could ever do because they can compress the original information in a sound signal or a picture. But the more you compress it, the worse the quality can get so you choose how much money you want to spend and what capacity you want to use . . . . ..

 

[Darwin123]

I was unable to find an exact figure, or anything that directly states this, but I found from a little reading, that there are numerous cables connecting various locations. So it's not just one cable carrying everyone's information to and from other continents. It is several, otherwise, any cable damage(which aren't uncommon from what I've read), be it natural, accidental, or even purposeful, could severely cripple international communications.

The wiki article for "Submarine communications cable" was quite interesting. http://en.wikipedia.org/wiki/Submarine_communications_cable There's even talk of espionage during the cold war! Fascinating stuff!

It not the speed of light that makes optic communication so fast. It is the frequency of the light.

The frequency of light is millions of times larger than the frequency of the signals in a conducting cable. Metal cables have a high frequency cut off. That is why metal cables look opaque. The optical allows through high frequencies. Not only higher frequencies, but more of them. The bandwidth of an optical cable is millions of times larger than the bandwidth of a metal cable the same diameter.

The speed of light and the speed of an electrical system aren't that different. The changes in electric current transmit through a copper cable roughly one third as fast as light pulses in a vacuum. If light in a vacuum travels 186,000 miles per second, an electrical pulse in a copper cable travels roughly 30,000 miles an second. The light pulses in a fiber optic travel roughly 2/3 the speed of light in a vacuum, or 60,000 miles a second. The speed of transmission isn't that different.

Frequency bandwidth [cycles per second]≈rate of information transfer [bits per second].

The frequency bandwidth determines the maximum number of Megabytes per second. The number of Megabytes per second for an optical cable numbers in the quadrillions. Each channel of communication takes a certain share of band width.

Different types of multiplexing are used to divide the available bandwidth up into channels. However, that is just a continuation of the bandwidth story.

 

[Drakkith]

Darwin, while you are pretty much correct, you do realize that the way signals are sent and received in the optical range is very very different than lower bands, right? We can't fully appreciate the higher wavelength of light because we rely on photodetectors to receive the light and generate an electrical signal. If we were able to directly measure the EM wave amplitude as it changes, which is what happens in normal transmission lines and antennas, we could further increase the transmission rate. (Assuming computer systems were able to keep up of course)

 

[davenn]

... The speed of light and the speed of an electrical system aren't that different. The changes in electric current transmit through a copper cable roughly one third as fast as light pulses in a vacuum. If light in a vacuum travels 186,000 miles per second, an electrical pulse in a copper cable travels roughly 30,000 miles an second. The light pulses in a fiber optic travel roughly 2/3 the speed of light in a vacuum, or 60,000 miles a second. The speed of transmission isn't that different. ...

and that's very wrong too.
the speed of an EM wave along a copper conductor is better than 95% the speed of light
and if an insulated conductor like a coax cable, its typically ~ 0.66 for a velocity factor.
Depending of the dielectric used foam, teflon, air etc the Vf of the cable can get up into the high 80's %

If all us radio techs went with your theory, our tuned lengths of coax and antennas would be very much different to what they really are ;)

Dave

 

[Omega0]

I would recommend to have a look at the OSI model (http://en.wikipedia.org/wiki/OSI_model). I did not have a deeper look at this article to tell you about the quality but the lower levels of the model should be understood first. I am pretty shocked when I read "The best known frameworks are the TCP/IP model and the OSI model." in the article http://en.wikipedia.org/wiki/Communications_protocol. I should post a note. The TCP/IP protocol are called that way from history but they are in two different layers. The more both layers are in the OSI model. Did I miss something?
What you will learn is how communication between system works (I guess not by this article but by a good book. I can't recommend one because the book I read was in German, it was brilliant and the author too).
I am not sure if only speaking about layer 1 does answer your question.

 

[Darwin123]

and that's very wrong too.
the speed of an EM wave along a copper conductor is better than 95% the speed of light
and if an insulated conductor like a coax cable, its typically ~ 0.66 for a velocity factor.
Depending of the dielectric used foam, teflon, air etc the Vf of the cable can get up into the high 80's %

If all us radio techs went with your theory, our tuned lengths of coax and antennas would be very much different to what they really are ;)

Dave

Oops. My mistake.
I was taking it from memory which is apparently faulty. However, you supported my main point. In terms of the speed of the signal, the copper conductor isn't much different from a fiber optic.

 

[Darwin123]

Darwin, while you are pretty much correct, you do realize that the way signals are sent and received in the optical range is very very different than lower bands, right? We can't fully appreciate the higher wavelength of light because we rely on photodetectors to receive the light and generate an electrical signal. If we were able to directly measure the EM wave amplitude as it changes, which is what happens in normal transmission lines and antennas, we could further increase the transmission rate. (Assuming computer systems were able to keep up of course)

That is why I specified frequency bandwidth rather than just the frequency. The amount of information is limited by the total bandwidth of the fiber optic. A fiber optic doesn't just carry one high frequency, it carriers a wide range of frequencies. If fiber didn't carry a wide range of frequencies, then the light traveling along it couldn't be modulated at a high frequency.

The frequency of the "carrier waves" and "sidebands" in a fiber have to be smaller than the optical frequency. Hence, the higher the optical frequency carried by a wire the larger the hypothetical limit on the rate of information transmission.