Historical BT paper on VDSL (1)

Found this on Ofcom's site. It shows the state of BT's thinking on VDSL (1) in 2001: link to Ofcom

At the time, the thinking was that VDSL would get you 14Mbps downstream, and 3Mbps upstream, using up to 12MHz frequencies.

Deployment would either be by the extra cabinets we now see, or by "bricks" in the footway boxes.

I've got a paper copy of that somewhere ...

Why has it taken so long to be deployed? OFCOM and equipment availability.

In reply to a post by MHC:

Why has it taken so long to be deployed? OFCOM and equipment availability.

Other probable reasons are cost and demand.

Back in 2001, ADSL was in the very early stages of roll-out (our exchange, serving 50,000+ people, had only recently been enabled for ADSL1 when this report was written).
Nobody really knew how many people would go for ADSL1 (then priced at around £40/month for 512kbps).
Most people were still on dial-up, or didn't even have an internet connection.

With over 80,000 cabinets (nationwide), the cost is tremendous.
If you assume, maybe £10k per VDSL cabinet (probably not too far off, when you consider the cost of the enclosure, the equipment inside, power hook-up, installation and commissioning), you're talking around £800million for the cabinets alone (could even be more).
Then you've go to add on the cost of installing maybe a quarter of a million km of fibre to serve these cabinets.
Then you've got to add on exchange equipment and back-bone infrastructure.

It's not really a surprise it's taken this long for a partial roll-out.

Also BT weren't too far behind other countries (most started rolling it out around 2008-2010). The big surprise was Russia, which started a roll-out in 2005!

Several other factors:
- The standards were also evolving rapidly; ADSL2 was able to surpass the early expectations of VDSL.
- The bulk of first generation roll-out continued for some time; I didn't see it until 2003.
- As Ade says, demand is a big factor; only now are services like streaming HDTV appearing to exploit
this bandwidth. OK, it's chicken and egg, but ADSL has been good enough for most people; now they want
more.

It'll be interesting to see how far VDSL2 can go in putting-off large-scale FTTP. The 17 MHz profile is an easy win but what next? Enhancements such as G.Vector could reduce the cross-talk issues and extend the reach of ~80 Mbit/s profiles (see, e.g., this multi-vendor paper.)

Living in a semi-rural area in Wales, I never expected to see 20 Mbit/s or more before ~2015; as it is, the little green cabinet at the end of the lane is half installed and, if not ready for the predicted Sep. 30, might be by the end of the year.

I meant to add that I thought I'd seen a figure in excess of £10 billion for the planned roll-out of FTTC/FTTP by ~2015; anyone else recall that?

£2.5bn is the number in my head

In reply to a post by yarwell:

£2.5bn is the number in my head

I seem to remember c. £2.5bn for the FTTC roll-out and anything up to around £15bn being quoted for a full nationwide FTTP roll-out.
Not difficult to see why BT went down the FTTC route. Especially considering it's relatively future-proof.

Fortunately the people who said I was wrong (on here, a few years back) when I suggested maybe BT could locate DSLAMs into street cabinets and fibre from the exchange to the cabinets (thus making all line lengths significantly shorter) turned out to be wrong themselves.
I remember "issues" (reasons for not doing it) being quoted as ventilation, power, fibre costs, etc.
It was nice to be right for once

In reply to a post by DLS:

Enhancements such as G.Vector could reduce the cross-talk issues and extend the reach of ~80 Mbit/s profiles (see, e.g., this multi-vendor paper.)

Wow, that's impressive stuff - particularly impressive is the way that it makes the speed much more predictable than the "plain" 17a profile... The figure for a 600m line suggests a range of 50-65mbps with plain 17a, or 95Mbps for the vectorised setup.

I guess it isn't quite the no-brainer that 17a is: it is likely to still need new end-user modems, and possibly new line cards in the FTTC cabinets (new firmware at the very least).

With FTTC capability like that up it's sleeve, it is going to make the FTTC model last at least a decade - and might be enough to get BT to consider adding some new cabinets to cover the longest lines out from the existing cabinet setup.

I just read the paper in a little more detail - though it's still a fair way over my head. But it does seem to show one thing - that all the work, both upstream & downstream, is done in the DSLAM.

That sounds like there is potential for this to be another upgrade that is done in the cabinet only!