Transfer *several* times as much data without increasing the size of the line

[TheScienceOrca]

Hey everyone, so for fiber optic networks we can either (in analogy) increase how fast the cars or photons move or add more lanes to increase the amount of data.

What if now we use the color of the light to interpret several times more data?

 

[davenn]

Hey everyone, so for fiber optic networks we can either (in analogy) increase how fast the cars or photons move or add more lanes to increase the amount of data.

'
that's a really bad analogy ... the speed of photons in a fibre is constant

What if now we use the colour of the light to interpret several times more data?

different wavelengths are already used ... but there is a dozen different modulation types that are more commonly used

do some googling on digital modulation schemes cdma, tdm, qam etc etc
they are used for radio links as well eg mobile phone service

Dave

 

[Baluncore]

increase how fast the cars or photons move

The photons travel at the speed of light in the fibre, that is fixed. Increasing the bandwidth of the optical modulator and demodulator increase the available data rate.

Using several different colours of light is already done to get greater data rate over fibre by having multiple broadband carriers. That is the same concept as having radio stations on different frequencies and in different bands.

 

[davenn]

thanks for the backup baluncore :)

 

[sophiecentaur]

This is a really vast can of worms.
There are two other major factors in the rate at which data can be transmitted are how fast you can switch the light on and off (or change its level), and detect this change at the receiver. Also, the amount of loss in your transmission chain (signal to noise ratio at the far end) will govern how easy it is to detect what data is actually being sent. Slower systems can be more robust.
The slight overall delay between transmission and reception is not usually a problem.

 

[Windadct]

IMO - Data compression trumps them all...or as I typically refer to it "really wacky math that is way beyond my comprehension"

 

[sophiecentaur]

IMO - Data compression trumps them all...or as I typically refer to it "really wacky math that is way beyond my comprehension"

The Shannon limit beckons . . . . . . .

 

[jim mcnamara]

There are network protocols for which compression on fast connections actually slows down transmission rate - ssh is one. (scp, ssh, sftp...).
Read a manual page for ssh where it clearly states this:

-C' Requests compression of all data (including stdin, stdout, stderr, and data for forwarded X11 and TCP connections). The compression algorithm is the same used by gzip(1), and the ''level'' can be controlled by the CompressionLevel option for protocol version 1. Compression is desirable on modem lines and other slow connections, but will only slow down things on fast networks. The default value can be set on a host-by-host basis in the configuration files; see the Compression option.

So, no, compression is not the only answer - and sophiecentaur's comment about dredging up worms applies. This thread has the potential to degenerate into an opinion fest. Which is not what PF is all about.

 

[dlgoff]

This is a really vast can of worms.

Indeed.

... light sources sometimes used with multi-mode fiber produce a range of wavelengths and these each propagate at different speeds.

http://en.wikipedia.org/wiki/Multi-mode_optical_fiber

Single mode fibers are therefore better at retaining the fidelity of each light pulse over longer distances than multi-mode fibers. For these reasons, single-mode fibers can have a higher bandwidth than multi-mode fibers.

http://en.wikipedia.org/wiki/Single-mode_optical_fiber

 

[Windadct]

Points above are valid - I think it really depends on the scope of the OP, the compression applied in the channel(Tx-chan-Rx) has its limits... but the DATA can be compressed / processed ( basic MP3 is about 10:1) - and "noisy" but can then be post processed(this is where the Math I mention comes in). Data(raw message) -Compressed -Tx -Channel-Rx- Decompressed - Data(message) ... EEs ( I am one ) are in love with fidelity-we like our systems to be "perfect"... but the world sometimes does not care or need 100% perfection - and no matter how hard you try there is no perfect communication channel... like debate in a forum ... :D

 

[analogdesign]

Multi-mode fiber is used because it is cheaper. The fact that it contains multiple propagation modes is not a good feature, it's a bug. Sophisticated algorithms and hardware are required to equalize the dispersion caused by multi-mode fiber. Similar techniques can extend the reach of the chromatic dispersion that limits the range of single mode fiber, but typically modal dispersion in multi-mode fiber is much, much more troublesome.

http://www-inst.eecs.berkeley.edu/~ee233/sp06/student_presentations/EE233_EDC.pdf

 

[sophiecentaur]

I think this thread has got well away from the simple requirements of the OP.
1. The motor car analogy is not very helpful - except if you tried to send a signal by writing numbers of letters on the sides of passing vehicles. The vehicle speed is fixed (motorway limits correspond to c on the fibre) but you could increase the data rate by using shorter vehicles - Smart cars rather than 18 wheelers. Thats like sending shorter pulses of light down the fibre so you can send more per second.
2. The use of a number of different "colours" of light down the same fibre is a well known technique and it's called Wavelength Division Multiplexing.

 

[TheScienceOrca]

Of course the photons travel the same speed I never said the latency would increase. I said if you use multiple wavelengths and interpret these wavelengths you could have more transfer which I guess they already do.