fiber optic network cables

[stickybytes]

^^ read this from a cisco networking textbook. how is this possible?

 

[ZeroNine8]

Perhaps because there are theoretically an infinite number of wavelengths in the spectrum that data could be transmitted on simultaneously? Just my guess.

 

[everman]

It's called marketing.

 

[imgod2u]

Many reasons, all of them theoretically plausible but not really applicable. As someone mentioned, the color spectrum can be split into an infinite number of spectrums to transmit data. Another reason is that fiber optic wires (if you can call it that) don't suffer from the same EM interference that electric wires do and thus, you can have tons of wires in parallel and they would interfere with each other to a relatively small (if at all) degree. Problems with EMI is a big problem with transmitting over large distances.

 

[Fencer128]

Hi,

In terms of bandwidth - it's a few terahertz (so definitely not infinite) if all of the available confined spectrum can be utilitsed (assuming SiO fibres with OH- absorption peak fully removed). That is to say radiation that can be successfully confined within the core.

Cheers,

Andy

 

[Skykat]

The reason they say that is because normal copper network cables have signal degradation when they try to send signals at very high speeds. the cables need to be of higher and higher quality to reduce the signal degradation to have an useable signal. Fiber optic cable does not have this problem, so, theoretically, fiber optic cable has infinite bandiwdth as long as you can create a device that can switch the light on and off faster and faster.

 

[TuxDave]

I would agree on the marketing answer. The bandwidth of fiber optics is definitely not infinitely wide. However, here are my two possibilities.

1) They may be referring to spectral bandwidth (in existance, not what can be transferred in a fiber optic cable), and that has no boundary that I know of.

2) Maybe it's like those 3000 night time minutes. It's far more than you can use so we call it virtually unlimited. Perhaps in this case, we are not limited by the bandwidth of the fiber optic cable but limited by the optical amplifiers, transmitters and recievers. Hence, the bandwidth is 'virtually unlimited'.

 

[iwantanewcomputer]

my school has fiber optic cables. you have to get a netgear ga621 fiber gigabit card ($200 new) or some giant converter switch($600 new) for laptops. in the end, campus-wide gigabit network and gigabit internet connection are worth it

 

[Fencer128]

Originally posted by: TuxDave
I would agree on the marketing answer. The bandwidth of fiber optics is definitely not infinitely wide. However, here are my two possibilities.

1) They may be referring to spectral bandwidth (in existance, not what can be transferred in a fiber optic cable), and that has no boundary that I know of.

2) Maybe it's like those 3000 night time minutes. It's far more than you can use so we call it virtually unlimited. Perhaps in this case, we are not limited by the bandwidth of the fiber optic cable but limited by the optical amplifiers, transmitters and recievers. Hence, the bandwidth is 'virtually unlimited'.

Hi,

The spectral bandwidth is around a micron in OH- absorption free Si fibres. The bandwidth (in terms of data throughput) is up around a few terahertz.

Cheers,

Andy

 

[TuxDave]

Originally posted by: Fencer128

Originally posted by: TuxDave
I would agree on the marketing answer. The bandwidth of fiber optics is definitely not infinitely wide. However, here are my two possibilities.

1) They may be referring to spectral bandwidth (in existance, not what can be transferred in a fiber optic cable), and that has no boundary that I know of.

2) Maybe it's like those 3000 night time minutes. It's far more than you can use so we call it virtually unlimited. Perhaps in this case, we are not limited by the bandwidth of the fiber optic cable but limited by the optical amplifiers, transmitters and recievers. Hence, the bandwidth is 'virtually unlimited'.

Hi,

The spectral bandwidth is around a micron in OH- absorption free Si fibres. The bandwidth (in terms of data throughput) is up around a few terahertz.

Cheers,

Andy

Umm... ok.... but read what I wrote again. I'll bold the important parts.

1) I was saying there was no bound of the spectrum of light that I know of.
2) I introduce the term 'virtually unlimited'

I know that the bandwidth in fiber is limited.

 

[Fencer128]

Originally posted by: TuxDave

Originally posted by: Fencer128

Originally posted by: TuxDave
I would agree on the marketing answer. The bandwidth of fiber optics is definitely not infinitely wide. However, here are my two possibilities.

1) They may be referring to spectral bandwidth (in existance, not what can be transferred in a fiber optic cable), and that has no boundary that I know of.

2) Maybe it's like those 3000 night time minutes. It's far more than you can use so we call it virtually unlimited. Perhaps in this case, we are not limited by the bandwidth of the fiber optic cable but limited by the optical amplifiers, transmitters and recievers. Hence, the bandwidth is 'virtually unlimited'.

Hi,

The spectral bandwidth is around a micron in OH- absorption free Si fibres. The bandwidth (in terms of data throughput) is up around a few terahertz.

Cheers,

Andy

Umm... ok.... but read what I wrote again. I'll bold the important parts.

1) I was saying there was no bound of the spectrum of light that I know of.
2) I introduce the term 'virtually unlimited'

I know that the bandwidth in fiber is limited.

Hi,

Sorry - I wasn't criticizing - I was just in a rush so I got the info down quick. By spectral bandwidth I mean the useable wavelength range over which there is ~0.2dm/km loss or less (which is practical for most comms apps). This is ~ a micron or so in modern fibre.

By bandwidth (data throughput) I mean the max data rate/frequency for a signal/wave that is not wavelength division multiplexed in any way (operating as quasi-monochromatic), which is about a few terahertz or so. This is a material property of the fibre and not the network components.

Hope that's cleared things up.

Cheers,

Andy

 

[Mik3y]

Originally posted by: everman
It's called marketing.

hahaha, that's the best damn answer ever.

 

[JKolstad]

Originally posted by: Skykat
The reason they say that is because normal copper network cables have signal degradation when they try to send signals at very high speeds. the cables need to be of higher and higher quality to reduce the signal degradation to have an useable signal. Fiber optic cable does not have this problem

Balogna. Fiber tends to suffer from pretty much all the same problems that any other transmission medium does when it comes to transmitting signals, it's just that you're starting with a carrier and materials that 'perform' up in the terahertz range so those problems don't really become significant until you start talking about bandwidths that -- compared to what you can typically do with copy -- are huge. It also has a couple of its own problems that don't usually crop up with, say, coax, such as modal dispersion. (It's not an uncommon question for undergraduate students to work out the frequency of the first non-TEM mode for coax cable... it comes as something of a surprise to many people that it's often at frequencies you standard a chance of hitting with contempotary signal generators.)

so, theoretically, fiber optic cable has infinite bandiwdth as long as you can create a device that can switch the light on and off faster and faster.

Don't be so naive, and don't post unless you know what you're talking about!

 

[cquark]

Originally posted by: ZeroNine8
Perhaps because there are theoretically an infinite number of wavelengths in the spectrum that data could be transmitted on simultaneously? Just my guess.

True, but optical fiber isn't transparent to all wavelengths and, of course, if you increase the frequency enough, you'll get gamma rays that will gradually destroy the fiber. Then there's the problem that it requires an infinite amount of energy to generate such an infinite spectrum.

 

[blahblah99]

Bandwidth through the fiber optics is unlimited, but the speed of the electronics will be limited (ie, transistors). I think the fastest transcievers being developed right now is somewhere in the tens of gigahertz.

 

[TuxDave]

Originally posted by: blahblah99
Bandwidth through the fiber optics is unlimited, but the speed of the electronics will be limited (ie, transistors). I think the fastest transcievers being developed right now is somewhere in the tens of gigahertz.

Please read the thread.

 

[jolancer]

i dont know any of the technical aspects of the optic network, except for what some have just stated, but... setting the optic cable out of the equation.. its safe to assume there is no such thing as infinate bandwidth, cause u would still be limited to the speed of light, right? [<- lmao, can u imagine an online gamer complaining about that bandwidth a few hundred thousand years from now]

 

[wirelessenabled]

Not much in this thread on dispersion problems with fiber. Bandwidth is certainly not unlimited.

Dispersion

As I recall much of the effort in increasing bandwidth for fiber has been expended on dispersion problems. Optical amplifiers, different frequency lasers, higher power lasers, have all been brought to bear on this. More advanced algorithms are also helping to sort out the wheat from the chaff.

 

[ZeroNine8]

The speed of light is only a limitation on the bandwidth at one wavelength. A single fiber can carry multiple wavelengths, so the ultimate limitation is the number of different wavelengths that can be transmitted simultaneously.

 

[BespinReactorShaft]

Here's my take based on what's in use today in the industry.

Sure, infinite number of wavelengths can exist in a single fiber. It's not practically true that the usable spectrum width is infinite; optical fibers carry light towards the infrared part of the spectrum, which in turn is divided into "bands." Each band is subdivided into a "grid" of individual wavelengths according to ITU-T standards. Three major bands are defined (the width and placement of each band is chosen because they deliver optimal attenuation behaviour; in other words the chemical composition of fiber is such that light signals come through stronger only in certain parts of the spectrum). So far three "bands" (S,L,C) are defined, although the most used are the C and L bands. Each band has 80 wavelengths defined, the spacing between wavelengths is 50GHz.

Reason for this standardised partitioning of wavelengths is, a nasty set of optical phenomena known as "non-linearities" creep in once you pack in too many wavelengths such that the spacing between wavelengths becomes too close. A major issue concerns optical cross-talk when wavelengths interfere with each other. The use of multi wavelength optical transmission is known better as WDM (wavelength division multiplexing) in industry circles.

Dispersion has always been a PITA for optical transmission, as it tends to limit the maximum transmission distance. At the same time it gets worse as the bitrate goes up, which is now a major barrier against long haul 40Gbps systems. The traditional way of combating dispersion has always been via dispersion compensation modules. In the past there have been creative efforts to design fibers with certain refractive index profiles such that dispersion is minimised or eliminated at certain wavelengths. But killing dispersion brings the non-linearity back up, which means a workable solution can't have the best of both worlds.

So far the most widely marketed field-ready WDM solutions has been around 160 wavelengths at 10Gbps per wavelength which would deliver 1.6Tbps on a single fiber. Alternatively there are labs which have demo'd 80 wavelenghts @ 40Gbps each. I'd be interested to hear of any "bleeding edge" capacity that exists today which is higher. Some engineering teams are already working towards 25GHz spacing, which would essentially double the capacities mentioned above.

By the way, Cisco is not exactly at the forefront of multi wavelength optical transmission among today's big-boy companies. You'll find better minds at work in companies like Alcatel, Marconi and (perhaps in former times) Lucent. Then again, Cisco has the biggest sounding trumpet out there.