How is data encoded in copper/light?

[KillaChaos]

I was watching the Silicon Photonics demo Intel posted (http://www.youtube.com/watch?v=vz3DaACN_54&feature=player_embedded) and I got interested in how data is stored in electrons/light. I previously thought that a "1" or "0" was sent to a chip depending on if the light was on or off. However this photonics demo looks different. He described muxing/demuxing process as taking a whole bunch of different wavelength of light, with data "in" them and sending them through a wire. I am guessing a different wave length (for example 400nm-405nm = 1 and 405nm-410nm = 0) would denote the data but I cannot find any articles that back up my claim for this process. Also, I realized that I have no idea how data travels over copper, nor do I know where to begin (I probably don't know much about electricity, but I don't know how you can have a signal in a copper wire when electrons are just flowing).

I hope my questions are not too poorly worded and if anyone can help me wrap me head around this it would be a huge help.

 

[berkeman]

I was watching the Silicon Photonics demo Intel posted (http://www.youtube.com/watch?v=vz3DaACN_54&feature=player_embedded) and I got interested in how data is stored in electrons/light. I previously thought that a "1" or "0" was sent to a chip depending on if the light was on or off. However this photonics demo looks different. He described muxing/demuxing process as taking a whole bunch of different wavelength of light, with data "in" them and sending them through a wire. I am guessing a different wave length (for example 400nm-405nm = 1 and 405nm-410nm = 0) would denote the data but I cannot find any articles that back up my claim for this process. Also, I realized that I have no idea how data travels over copper, nor do I know where to begin (I probably don't know much about electricity, but I don't know how you can have a signal in a copper wire when electrons are just flowing).

I hope my questions are not too poorly worded and if anyone can help me wrap me head around this it would be a huge help.

A keyword is "modulation" -- that is how information is encoded into the carrier medium:

http://en.wikipedia.org/wiki/Modulation

.

 

[waht]

] Also, I realized that I have no idea how data travels over copper, nor do I know where to begin (I probably don't know much about electricity, but I don't know how you can have a signal in a copper wire when electrons are just flowing).

I hope my questions are not too poorly worded and if anyone can help me wrap me head around this it would be a huge help.

It's not just "electrons" that are flowing, rather than, a whole bunch a independent radio waves is traveling which can exist at different frequencies or channels within a copper/fiber medium. The differences between copper and fiber lies in the number of radiowaves you can squeeze in. In optical fibers, you can squeeze in trillions more waves layman speaking than in copper.

So now that you have a radio wave riding in a medium, let it be copper or fiber or whatever, how do you send data?

The answer lies in modulation as berkeman stated, and there is many different types. A simple one is amplitude modulation. Simply represent a strong wave with 1, and a weak wave with 0, and send them along.

 

[Mike_In_Plano]

Typically with fibers, the intensity of the light corresponds to a 1 or 0 and the intensity of the light can be varied rapidly - billions of times for second. However, this doesn't make the best use out of the fiber, because there are different colors or "wavelengths" of light that can all be passing through the same fiber. Each wavelength starts from a separate laser diode, and and now each laser diode can be controlled with 1's and 0's.

Thus you can send 10 channels all down the same fiber (10 x the communications bandwidth) by using 10 separate lasers.

On the other end, the wavelengths need to be separated, detected, and converted back to electrical signals. Again, 10 colors must be separated into 10 wavelengths and in turn into 10 multi-gigabyte signals.

This is pretty exciting stuff, but it can get tricky. The lasers change wavelength as their temperature changes. Thus they need temperature controls. Also, the color filters need to be carefully designed and fabricated. It's even possible, at the crazy limits of the technology, that the bandwidth of the color filter can interfere with the bandwidth of the recovered electrical signal because such a narrow band is filtered through that it "averages out" the passed signal!