Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Bowdon]

If a fibre line to the cabinet had for example SNR DS 30dB does that maintain to the cabinet, and then a drop off happens while it uses the copper lines to reach the end user?

I'm wondering because if it does start off at 30dB and a person has FTTP technology installed. Would they get the 30dB at the end user?

I'm assuming the end users house has perfect internal connections.

[GigabitEthernet]

Fibre works differently because data is transferred by light. I believe what is measured is light loss which is very, very low over any distance so in effect the light would be as strong at the cabinet as it did when it started at the exchange.

[Weaver]

I think it's fair to say that with fibre optic comms including FTTP there's no such thing as SNR in the sense you mean it, the sense that it is used in with DSL.

It's simplest to think of FTTP as a system that just works, it gets the data there like comms over your Ethernet LAN.

[Weaver]

But at a physical level as opposed to the usual comms viewpoint adopted in DSL, as a previous poster said, the light level down the fibre decreases with distance and the signal-to-noise level therefore goes down. This optical / physical SNR isn't going to necessarily be visible to the end-user though.

In DSL, the ‘SNR margin’ figures quoted are only very indirectly connected to 'physical SNR', connected via a ton of complex maths and through the details of the signal decoding algorithm. When I made up the phrase physical SNR back there, I meant the signal amplitude in Volts or Watts divided by the noise amplitude, akin to the sort of definition used in physics.

[GigabitEthernet]

Weaver, would it even be called Signal to NOISE ratio as light isn't affected by noise?

Whatever it's called (and I don't know), it would make effectively no difference because light travels such a long distance. This is why you can be, in effect, as far away from the exchange as you want when using pure fibre as the light level doesn't change enough to make an impact.

[Weaver]

@AlecR - physicists would be happy to call the ratio of powers or ratio of amplitudes SNR. light does not travel down fibre optics for tens of kilometres without impairment, although dispersion which is a big performance killer does not get counted under the heading of SNR. There will always be a noise level, in the sense that the optical receiver will be picking up some kind of optical noise signal in the event of no data signal. (That noise could be leakage or thermal noise, or who knows what.)

So I have no problem using the term SNR. but that doesn't answer your question. I suppose it depends on who you ask, as to whether or not the SNR concept feels like a good fit from a particular vewpoint.

[Weaver]

Light is an electromagnetic phenomenon, just a really high frequency one. You are using the term noise in an unreasonably restricted sense to mean only low-frequency electric field or magnetic field disturbers.

Noise in acoustics is just, well, noise. There are lots of differentiate kinds of application where the term noise is used.

It's like flowers and weeds. Mathematically noise is the weeds, signal is the flowers.

[GigabitEthernet]

Weaver is your expert here, I am just a mere mortal

All I know is that you can essentially be as far away from the exchange as you want and pay for a 330Mbps line from Openreach (or you could before they suspended FTTPoD) but it will cost you.

[guest]

Never really heard of SNR for monomode/multimode/etc fibre.

Dispersion is more appropriate.

[Bowdon]

So it really does seem like fibre is the next big step in technology to the old copper lines.

I think the quicker we're on to full fibre, whether that be FTTP or via a G.fast 2-step, it'll be a good thing.

It would seem also that with interference being less of a problem, I bet the number of engineer call-outs will drop also.

[kitz]

Rizla beat me to it and Weaver is correct when he says "there's no such thing as SNR in the sense you mean it,"
SNR in this sense relates to copper which carry (high frequency) Digital Signals which can be impacted by Noise.  With DSL its all about Signal strength and Noise hence SNR measurement being important.

With fibre, the signals are converted into pulses of light - See here for how its done.

Light isnt subject to interference from Noise, but it can spread. Crude example:-
 
Think of being stood at one end of a football pitch and your friend at the other.   He can shout but his voice will be less heard the further he is away.  If there's a crowd jeering, that introduces noise, so you stand even less chance of hearing him shout.  Thats SNR.
Compare to light.  Your friend holds a torch, but the beam can spread (disperse), so the trick is to contain the light source in cable where it cant disperse or be affected by other light sources (such as the sun).
 
   
Light dispersion is an issue with Multimode fibre.  Not so with Single mode fibre, but that is expensive and usually only used for backhauls.   Wave Division Multiplexing is one way of getting around transmitting multiple signals on 'single mode' fibre but again that is relatively expensive, but is used extensively by BTw in their Core and backhaul as its cheaper than laying more fibre.

Presently WDM isnt used much in the local loop by BTw, but they do have plans to do so with FTTP and FTTdp.     

[Weaver]

Referring to Kitz' use of the word dispersion. In physics, dispersion means that the speed of a wave may vary according to frequency/wavelength if the wave is travelling through a certain type of medium. In fibre optic cables, light can end up spreading out in the direction of travel, so a transmitted pulse ends up becoming longer front-to-back. This lengthening of a pulse is performance killer because the receiver gets blurry lengthened pulses which are the wrong time length and ultimately could blur together in time. This is why a dispersive glass has speed performance limits. Some physicists are now using hollow glass pipes, where the light travels through the empty space insider the pipe, not through the glass.

If Kitz were talking about a beam of light in air spreading out sideways, which I doubt she is, this in itself is not covered by the term dispersion, for example, a torch beam spreading out sideways because the beam edges are not parallel and the width of the beam increases with distance, this is not an example of “dispersion” as used in physics.

When optical dispersion happens due to the properties of a particular transmissive material, light can be spread out sideways when it is refracted by a change of medium and its path is bent if the medium it is entering is dispersive. An example is light entering glass or water. Dispersive mediums exhibit frequency-dependent bending angle. A glass prism shows this refractive bending and also sideways spreading due to dispersion in the glass, and white light separates out sideways into the familiar spectrum of constituent colours, as demonstrated by Newton.

[c6em]

@Kitz
Gigaclear are using single mode fibre throughout their installations right up into the house.
Here is a cable code spec from one of their multi-fibre-cored armoured cables (144 fibre count) which run along the roads to one of their amalgamation/splitter points on each road.
ACE-TKF (that's the manufacturer)
LTC-S CS 144 * SM G.657.A1(12*12)

Pretty sure B4RN are also using single mode throughout
I've no idea what BT are using for their GPON FTTP installs

[JGO]

I think there is a confusion between fibre and fibre system.

 The system needs a light source, and there is a loss because all the emitted light doesn't enter the fibre. This is worse for single fibre than multimode because of it smaller diameter.
 
When I was dealing with fibre systems the manufacturer's aim was to reduce the single fibre transmission loss to less than 1 dB/km; I recall a present day figure of 0.5 dB/km. Multimode was higher.

At the receiving end the aim is to capture as much as possible ( loss !)  of the light on some sort of light to voltage converter  and (unless it is at - 273 deg K !) this generates noise ! Think of a digital camera, or a film camera for that matter.

Thus there are three components of loss which reduce the light signal relative to the fixed receiving end electrical noise i.e. degrade S/N. Good system design would  use a method of modulation which is not (very) sensitive to input S/N over the expected range, like FM  radio which works with negligible audible noise over a wide range of S/N ( but captures on noise outside it !)