Kitz ADSL Broadband Information
adsl spacer  
Support this site
Home Broadband ISPs Tech Routers Wiki Forum
 
     
   Compare ISP   Rate your ISP
   Glossary   Glossary
 
Please login or register.

Login with username, password and session length
Advanced search  

News:

Pages: 1 ... 5 6 [7]

Author Topic: G.FAST  (Read 18280 times)

WWWombat

  • Kitizen
  • ****
  • Posts: 1674
Re: G.FAST
« Reply #90 on: April 09, 2015, 08:41:22 PM »

Except of course for those who don't get anywhere close to the current maximum because they're already too far away from the FTTC cabinet, and those who may not get any upgrade from what they have until BT decides to do something substantial (vectoring, pair bonding, FTTdp, FTTP)

Using your "gas supply" analogy, those "substantial" solutions are the exact equivalent of using tanks or bottles for gas. In both cases, it is about adjusting the business case - the finances - until a solution fits in well enough.

In the gas-supply case, pipes are used when the economics allow for it, but tanks and bottles are used when the economics don't pan out. They certainly can't do *everything* that mains gas can - the safety equation is very, very different, and the risk of supply failure is different. In the end, "gas supply" is the same - a pared-down, "just about good enough" bodge.

Quote
Everyone's situation is different, but around here everyone is either on overhead wiring or it's a new build with underground wiring, neat ducting and whatever. The latter already is FTTP. and so is some of the former

I guess you picked out the easiest, cheapest, two cases there.

Differences appear when the full cable is distributed overhead, rather than just the drop wire. The load on the poles needs to be considered - so fibre cannot just be added where a pole is already fully-loaded. Other utilities play a part too; it appears that current regulations means no more shared poles, and increased spacing from overhead electrical wiring too: what used to be done can no longer be done.

For underground, the issue comes back to ducting, and the question about whether there is spare capacity, and whether there has been any collapse - or even if ducting exists at all.

Then we feed into the next point: scale. And scale is a thing that doesn't enter most people's thinking.

Every argument you make is about your house, your line, your pole, and what amount of "imperfection" you are willing to put up with to get gigabit (by which I mean, you will accept extra cables outside, holes in walls, untidy skirting boards, the extra battery backup unit etc).

BT have to think at a scale way beyond just you: They have to consider a solution for everyone in an area; they have to cope with some streets having poles, some using poles shared with electricity, some being ducted, and some being direct-burial. They have to figure the total works required, and factor in things like main roads with traffic measures.  Whatever solution is picked has to balance over everyone - because some will be cheap to do, and some will be expensive.

Within this environment, they have to consider all the routes out from a cabinet, and all the junctions. Is there room in every single path to add fibre? Are the chambers large enough to add the extra underground equipment, or can enough space be found in the pathway to put a new chamber without disturbing other utilities?

Once you've factored the above considerations for individual routes, and individual homes, you have to add some coherency. If half of the cabinet area could get FTTP economically, and half could not ... which options do you go for? Do FTTP for half, and leave the others out? Or upgrade the cabinet to FTTC for everyone ... a decision which then destroys the economics for FTTP where it would have been viable?

Eventually, it leads to a decision: Cabinet for the whole area, or FTTP for the whole area.

There's an interesting Analysys Mason document on "Sample Survey of Ducts and Poles in the UK Access Network", linked on this Ofcom page. It gives a feel for the picture seen from a higher vantage point.

Beyond the technical and financial issues, there will be some emotional problems. Most people do not like having their prize garden dug, or their driveway dug; The likes of you and me, frequenting broadband forums, may be perfectly willing to compromise so we get fibre ... but most won't. Such issues either add to the cost, or add a barrier to take-up (making payback longer).

Quote
There will of course be situations where G.FAST is almost as good and doesn't have the same hassles as FTTP for certain areas with local infrastructure issues, I just hope it isn't picked as the one-size-fits-all solution like FTTC seems to have been

I think the nature of the current FTTC deployment makes it look like a one-size-fits-all solution, but it isn't. It is a one-size-fits-almost-everyone solution, combined with the current target levels (67% commercial, 90% subsidised superfast) that haven't yet required anything other than this one size. Even the next target level (95% subsidised superfast) can likely be achieved (on average, nationally) with just this one size solution too. But if you start setting 95% as standalone targets in each of the sparse counties, you are going to have to see the other solutions come out.

The nature of this "one-size" rollout is that it leaves the fringes of upgraded cabinets under-served. Around 4% suffer from this, judging by TBB stats, but that will be worse for some of the cabinets yet to be rolled out (the harder, rural ones).

It seems that FTTRN works for this 4%, in a technical sense, but suffers from the cost of power. We haven't seen what solution BT have for this, if anything.

Reverse power, for FTTdp, seems to offer a solution, but is probably a couple of years away. If they can get it working well, I can see it being rolled out in the way that FTTRN was trialled (Ulshaw, North Yorkshire).

FTTB also seems to be in the list for trials this year.

Right, but they do this for copper every day. An Openreach employee (appropriately insured) visited my house to install the FTTC modem. If I had line issues, someone would enter my house and twiddle with BT-owned internal wiring (which will have to be insured against the risk of it causing damage to my property). If I wanted more copper lines, again, more internal work, running new cable from the pole, drilling holes in an exterior wall, etc, all of which would be done with no trouble.

For most people, isn't the second line there already, in the existing cable? There is little extra cost that way, but not many homes use a second line nowadays - with the prevalence of mobiles and VoIP.

Quote
The large rural/urban difference in effort and cost is really going to be in getting the fibre from the headend to the pole, surely (which you'll be doing for G.FAST or FTTP anyway, and where a large percentage of the distance has likely already been covered with the FTTC build)

It depends. Overall, I think the total cost of getting to the pole (from the cabinet) about balances the cost of going from pole to home; actual amounts depend on the takeup. Obviously that means the cheapest home-pole cases will be worth doing as FTTP (providing that *every* home in the area can be done that way cheaply); the expensive home-pole cases will be worth doing as FTTdp.

The reason that the balance of cost is not as you think comes down, simply, to the opportunity for sharing the cost between homes.

- On average, the exchange-cabinet distance was about 2.8km, but the cost gets shared amongst 300 homes; that amounts to 9m of fibre per home; less if the fibre spines are shared significantly, which I suspect they are. They are also mostly ducted, helping costs.

- On average, the cabinet-DP distance is around 400m, and the cost gets shared amongst 8 homes; that amounts to 50m of fibre per home; less if the fibre sub-spines are shared significantly, which I suspect they are. They are also mostly ducted, helping costs.

- On average, drop wires are about 35m, and by definition, not shared.

Quote
- the additional few metres from pole to home is largely the same if it's overhead wiring in central London or overhead wiring in the Highlands, no? (if both houses were near their respective poles).

It really depends on how much overhead wiring there is in a location. I think that AM documents gives the national proportion of DP types, but how they are distributed is key.

I think, however, you are missing one or two elements that lie in G.fast's favour: The ability to roll out much faster, and the ability to do it without needing appointments with subscribers.
Logged

Black Sheep

  • Helpful
  • Addicted Kitizen
  • *
  • Posts: 5252
Re: G.FAST
« Reply #91 on: April 09, 2015, 08:55:41 PM »

A well constructed post (again) W3. The sticky-out salient point being BT/OR have to factor in every single hurdle that may confront them, before deciding on a business plan. That is why I continually try to get folk to see beyond their own little piece of the UK, which is usually a bog-standard dwelling.

You are absolutely correct about JUP's (Joint User Poles) not being an option on new provides. The Electric boards appear to honour existing wiring as being acceptable, but no new additions !! Then there's places of outstanding natural beauty, places of scientific interest, conservation areas, islands (Submarine cables) that have to be specifically planned around. I'm not a 'Planner' but I bet there's a thousand and one different scenario's that differ to providing service to a bog-standard dwelling ??

Logged

CarlT

  • Kitizen
  • ****
  • Posts: 1718
  • Random network numpty
Re: G.FAST
« Reply #92 on: April 09, 2015, 10:28:55 PM »

- On average, the cabinet-DP distance is around 400m, and the cost gets shared amongst 8 homes; that amounts to 50m of fibre per home; less if the fibre sub-spines are shared significantly, which I suspect they are.

G.fast will likely be backhauled via XGPON going forward, which does indeed provide creative options for reducing fibre count of backhaul.
Logged
WiFi: Nighthawk® AX12 RAX120
Routing: Mikrotik CCR2004-1G-12S+2XS
Switching: Mikrotik 2 * CRS305-1G-4S-IN, 2 * CRS309-1G-8S+; 1 * CSS326-24G-2S+RM
Exchange: Wakefield
ISP: BT Full Fibre 900. Zen Full Fibre 900. Faelix FTTP 300

sorc

  • Member
  • **
  • Posts: 28
Re: G.FAST
« Reply #93 on: April 10, 2015, 09:44:54 AM »

Using your "gas supply" analogy, those "substantial" solutions are the exact equivalent of using tanks or bottles for gas. In both cases, it is about adjusting the business case - the finances - until a solution fits in well enough.

In the gas-supply case, pipes are used when the economics allow for it, but tanks and bottles are used when the economics don't pan out. They certainly can't do *everything* that mains gas can - the safety equation is very, very different, and the risk of supply failure is different. In the end, "gas supply" is the same - a pared-down, "just about good enough" bodge.

Except that there are two differences here - the first being that the taxpayer hasn't paid billions for me to get a gas tank after promising mains gas (the home owner does this entirely out of their own pocket) - the local gas DNO is not involved, and the second being that there is no functional difference at all. I can have a gas cooker, gas fires, gas central heating. It works precisely the same as someone on mains gas, with the exception of having to find LPG appliances or ones that can be converted. Something that works 100% as well is not really a bodge.

That is not true for FTTC vs FTTP - both in attainable speed (most people on FTTC won't get 80Mbit, everyone on FTTP can) and in the future where that gap gets even worse.

I guess you picked out the easiest, cheapest, two cases there.

Differences appear when the full cable is distributed overhead, rather than just the drop wire. The load on the poles needs to be considered - so fibre cannot just be added where a pole is already fully-loaded. Other utilities play a part too; it appears that current regulations means no more shared poles, and increased spacing from overhead electrical wiring too: what used to be done can no longer be done.

For underground, the issue comes back to ducting, and the question about whether there is spare capacity, and whether there has been any collapse - or even if ducting exists at all.

Then we feed into the next point: scale. And scale is a thing that doesn't enter most people's thinking.

Every argument you make is about your house, your line, your pole, and what amount of "imperfection" you are willing to put up with to get gigabit (by which I mean, you will accept extra cables outside, holes in walls, untidy skirting boards, the extra battery backup unit etc).

BT have to think at a scale way beyond just you: They have to consider a solution for everyone in an area; they have to cope with some streets having poles, some using poles shared with electricity, some being ducted, and some being direct-burial. They have to figure the total works required, and factor in things like main roads with traffic measures.  Whatever solution is picked has to balance over everyone - because some will be cheap to do, and some will be expensive.

I picked it out because it isn't just the pole outside my house, it's pretty much how the entire local area is done.

I'm not sure if you appreciate how much FTTP is in places like Cornwall, and exactly where BT has decided to do it. The work to bury several km of fibre to serve an individual home in the middle of nowhere, for example - are you going to say that this is cheaper and easier than running an additional couple hundred metres of fibre from the aggregation node to the home in a more built up area and where costs can be split between more users? And as far as I can tell there is no real method or science to the type of technology they choose to deploy in these very rural areas. Hence some very rural premises getting FTTP and others getting told to go to satellite.


Within this environment, they have to consider all the routes out from a cabinet, and all the junctions. Is there room in every single path to add fibre? Are the chambers large enough to add the extra underground equipment, or can enough space be found in the pathway to put a new chamber without disturbing other utilities?

Once you've factored the above considerations for individual routes, and individual homes, you have to add some coherency. If half of the cabinet area could get FTTP economically, and half could not ... which options do you go for? Do FTTP for half, and leave the others out? Or upgrade the cabinet to FTTC for everyone ... a decision which then destroys the economics for FTTP where it would have been viable?

Eventually, it leads to a decision: Cabinet for the whole area, or FTTP for the whole area.

Most of your objections are going to largely apply for G.FAST - so it's a problem BT will need to tackle either now, or in the near future once it is realised that FTTC isn't cutting it. Especially where Virgin or some other higher-speed competitor exists.

Your argument also seems to omit the rollout costs of G.FAST - how much will the thousands of pole/pit mounted DSLAMs cost to install/maintain/replace per-user? Will it be cheaper than an equivalent FTTP install? (where it only really needs to be done once, and that's it). The upgrade path/futureproofability should have been considered too, not just how much it costs today.

So it's really a tossup between doing the work now, or doing it later - but after you've spent money on doing FTTC first.

Beyond the technical and financial issues, there will be some emotional problems. Most people do not like having their prize garden dug, or their driveway dug; The likes of you and me, frequenting broadband forums, may be perfectly willing to compromise so we get fibre ... but most won't. Such issues either add to the cost, or add a barrier to take-up (making payback longer).

I've already said that G.FAST might be more useful in certain situations - but I don't believe that applies to almost everyone - and notice that I have concentrated on overhead wiring where no such digging is needed.

I think the nature of the current FTTC deployment makes it look like a one-size-fits-all solution, but it isn't. It is a one-size-fits-almost-everyone solution, combined with the current target levels (67% commercial, 90% subsidised superfast) that haven't yet required anything other than this one size. Even the next target level (95% subsidised superfast) can likely be achieved (on average, nationally) with just this one size solution too. But if you start setting 95% as standalone targets in each of the sparse counties, you are going to have to see the other solutions come out.

Except that "superfast" in many cases seems to be speeds that were pedestrian 10 years ago, let alone 5 or 10 years into the future. As long as you can squeeze something out of the FTTC cabinet, that's "covered". Hence BT's decision to do ADSL2+ from FTTC in the hopes of getting a few more users for the precious "homes covered" statistic, even if the attainable speeds are utter pants.

For most people, isn't the second line there already, in the existing cable? There is little extra cost that way, but not many homes use a second line nowadays - with the prevalence of mobiles and VoIP.

I was referring to a situation where I'd want more phone lines than my existing dropwire can handle. BT would get that done with no problem (provided that there is suitable facilities to do it)


- On average, the exchange-cabinet distance was about 2.8km, but the cost gets shared amongst 300 homes; that amounts to 9m of fibre per home; less if the fibre spines are shared significantly, which I suspect they are. They are also mostly ducted, helping costs.

- On average, the cabinet-DP distance is around 400m, and the cost gets shared amongst 8 homes; that amounts to 50m of fibre per home; less if the fibre sub-spines are shared significantly, which I suspect they are. They are also mostly ducted, helping costs.

This is additionally a G.FAST problem too, is it not? You're going to need to get fibre to the pole/pit somehow - and then you plonk a G.FAST DSLAM on the end or a splitter and keep going into the home.


It really depends on how much overhead wiring there is in a location. I think that AM documents gives the national proportion of DP types, but how they are distributed is key.

I think, however, you are missing one or two elements that lie in G.fast's favour: The ability to roll out much faster, and the ability to do it without needing appointments with subscribers.

And you appear to be missing the actual cost of G.FAST here - DSLAMs and modems for what will be a brand new technology won't be cheap too (and it remains to be seen if G.FAST won't require a proper installation to get the best possible speed, since every metre seems to count, I doubt BT will want people plugging the things in at the end of 30 metres of B&Q special non-twisted pair extension cable)

And as for rollout speed - if FTTP rollout is allegedly slow (given what's happened locally I don't totally believe that), G.FAST is going to be the same. It might be easy for someone to come and nail a DSLAM to the pole, but it's still going to take a lot of time and expense to put that fibre backhaul in

At least the years of stagnation (8Mbit ADSL to most unless TalkTalk were in the exchange) on BT's part meant that VDSL2 was relatively mature and established before they decided to deploy it

A well constructed post (again) W3. The sticky-out salient point being BT/OR have to factor in every single hurdle that may confront them, before deciding on a business plan. That is why I continually try to get folk to see beyond their own little piece of the UK, which is usually a bog-standard dwelling.

But BT doesn't make these decisions at national level - there's already an element of choosing different technologies for a local/cabinet level. Hence some people having FTTP, most on FTTC, some having FTTrn, some being fobbed off onto satellite. So talking about a person's local area (in my case I was thinking more about parts of a county) is highly relevant.
« Last Edit: April 10, 2015, 10:20:02 AM by sorc »
Logged

guest

  • Guest
Re: G.FAST
« Reply #94 on: April 10, 2015, 10:16:49 PM »

Do you have any feel for whether Sckipio's recent announcement of faster G.fast, with longer range, stands up?

If it's the links you provided then in theory its possible. Its very unlikely in the field. 300Mbps over 300-400m - no problem, likewise 100Mbps over 1km isn't startling. What they're claiming is do-able but its unlikely to be the median result IMHO.

When they talk about bonding its important to understand that bonding two copper pairs with vectoring will produce an aggregate speed of more than double that of each pair. This is called "phantom mode" (google it). Essentially it means you have 3 DSL channels - one between the A/B wires on the first pair, one between A/B on the second pair and a third virtual channel between the pairs themselves. https://gigaom.com/2010/04/20/dsl-speed-300-mbps/ used to have a nice clickable graphic which showed this.

Edit - this expands beyond two pairs. Provided the pairs are in close physical proximity (which is likely for g.fast) then its basically the number of actual channels (N) plus N-1 virtual channels. So if you had 4 pairs then you'd have 4 actual + 3 virtual DSL channels. The maths on vectoring (with phantom modes) is horrendous & all sorts of analogue shortcuts are currently taken as the vectoring resources on the cheap as chips SoC aren't enough . Suffice it to say if you're a decent RF analogue engineer (rare as rocking horse **** in the UK now :( ) then you can more or less write your own ticket - if its contract & you're accepting under £100/hour then you need to find another agent ;)

tl;dr yes it probably stands up under limited conditions, the devil (as always) is in the detail.
« Last Edit: April 10, 2015, 10:39:20 PM by rizla »
Logged
Pages: 1 ... 5 6 [7]