Skip Navigation


Openreach G.fast roll-out to start in 2016/2017
Friday 30 January 2015 10:35:18 by Andrew Ferguson

We went at looked at some pre-release G.fast hardware at the BT test labs in 2014 and now three months later Openreach is announcing an ultrafast future with deployment starting in 2016/2017 some eight years after FTTC started to be rolled out.

G.fast is being described as delivering up to 500 Mbps and the BT ambition is to have this rolled out to most of the UK within a decade, so we presume 2025, in the shorter term the talk is of availability to millions by 2020. Of course this is subject to nothing untoward turning up in some public trials, and 4,000 homes and businesses in Huntingdon, Cambridgeshire and Gosforth, Newcastle will be able to take part in two pilot roll-outs in Summer 2015. This will build on the core work at Adastral park and help to train staff and verify that the expected speeds can be delivered.

Not everyone reading this will know what G.fast means, but basically it means a small box being attached to the pole that feeds your home, or dropped into a hole in the pavement and rather than your fibre broadband being delivered by a cabinet the fibre is ran to this small box and then your existing copper phone line delivers the G.fast signal. Higher frequencies are used which combined with the shorter distance means higher speeds are possible.

While we know that there will be those that complain that this is just Openreach exploiting its copper monopoly, the deployment of G.fast would not have been possible without the pushing of fibre out from the 5,500 exchanges to what is some 61,000+ street locations currently and pushing it again to a point within 100m or so of a cluster of premises means that just maybe in 2025 we may actually see the final decision to roll-out FTTP/H to millions by Openreach. Of course competitors like Hyperoptic, CityFibre and Gigaclear should have millions of FTTH connections by then too if their plans come to fruition.

What is clear is that the EU 2020 Digital Agenda target of 50% of the EU actually subscribing to a 100 Mbps or faster connection is likely to be a reality and the UK will be above the average across the EU. The lower standard of everyone with a 30 Mbps connection will depend on how the Governments final 5% pilots work out, though some areas are pushing ahead to 100% targets already.

The roll-out timing is interesting as it overlaps with the phase 2 BDUK dates and we would be shocked if G.fast did not feature in the superfast roll-outs in those local authorities aiming for 95% to 100% coverage at superfast speeds. The cost of powering a G.fast node may be the deciding factor between a small hamlet getting G.fast or FTTH/P.

Update 12:10pm In response to a few questions we have learnt that some G.fast nodes may be deployed from the fibre cabinets, the advantage being that DC power is available. Another alternative being DC power from the exchange. The need for mini-cabs and footway deployment is recognised and this announcement is about intention to deploy rather than a precise plan for the future. In terms of the future roll-out Openreach acknowledges that local demand hotspots will be one of the many factors and the trials will be looking closely at peoples desire to upgrade from superfast to ultrafast, i.e. how willing are people to pay for even more speed than they can sensibly use at once or put another way how valuable is waiting 1 minute for a large download versus 5 minutes.

A small note, there will also be a Gigabit trial, that we believe will look at delivering Gigabit speeds over the existing GPON GEA-FTTP network. Technically there is no problems, but understanding willingness of people to upgrade and pay for higher speeds and usage patterns are probably more a factor in that trial.

Comments

Posted by Spud2003 about 1 year ago
This article is dated 30th Jan but I'm pretty sure it's April 1st ... anyway this is a huge surprise for me.
Posted by ian72 about 1 year ago
Can't wait for a couple of years time when people start complaining that G.Fast has not yet been rolled out to them and so they can only get 80Mbps rather than the 500Mbps being rolled out to others. We seem to be doomed to a digital divide for the rest of eternity where some people will always be able to complain that "BT have left them behind".
Posted by SaticICE about 1 year ago
Also it is Adastral Park, not Ad Astral (I work on that very site :)
Posted by AndyCZ about 1 year ago
Nice to know that someone from TSO is on here :)
Posted by Kebabselector about 1 year ago
I'd be happy with just bog standard FTTC first.
Posted by jacko0 about 1 year ago
I agree Kebabselector.. FTTC would be nice if I could get it! Just upgrade the cabinet FFS!
Posted by TheEulerID about 1 year ago
I think the only cost effective way of deploying power to these nodes will be using reverse power. Delivering 3W per line that way would be 90% efficient over a 200m loop (at 48V). Just conceivably, a distributed network supply system could be engineered using the copper in the D side, but why do it?
Posted by andrew (Favicon staff member) about 1 year ago
@SaticICE Fixed

G.fast changes the economics and some cabinets that have not been enabled are for good reason, i.e. FTTC would not benefit the majority so little point currently.

For areas getting FTTrN (VDSL2) there is even the chance that a swap of hardware may boost speeds very easily.
Posted by TheEulerID about 1 year ago
Incidentally, North Yorks reports that the cost of getting power to FTTrN nodes is proving to be prohibitive. Some form of innovative RP or DP solution is required for that.


nb. if a node down a foot box is to support (say) 20 lines, the power budget will be limited anyway.
Posted by andrew (Favicon staff member) about 1 year ago
@TheEulerID see an update, power from cabs/exchange are solutions and even the G.fast located at the cabinet. Certainly will make trial, and getting deployment more widespread easier.

Obviously cabinets with larger footprints would still need remote nodes, but it is a start.

Posted by deanwade about 1 year ago
Fibre is cheaper than copper bit BT still installing copper to new developments.

This does not make sense... unless BT want to preserve their copper monopoly.
Posted by andrew (Favicon staff member) about 1 year ago
USO requirements and equivalence of input still require copper to be installed. Pure Fibre only developments via Openreach are problematic as Sky/TalkTalk cannot offer unbundled voice.

Is fibre cheaper to install than copper? Material costs can be different to actual install costs.
Posted by TheEulerID about 1 year ago
@andrew

Applying some basic electrical engineering principles makes me wonder about supply LT power over wire from cabinets.

Per pair (on standard gauge) the max deliverable at 48v is 8.3W at 200m, 3.4W at 500m and 1.7W at 1,000m (and that again is wasted).

Of course, dedicated (thick) cable can be used - expensive and bulky. Higher voltages can be used, but historically telcos don't allow that in shared ducts.
Posted by TheEulerID about 1 year ago
Also, reading the literature, g.fast does seem very much predicated on reverse power. There are consolidation circuits which draw a certain amount of power over several pairs to provide 12V for nodes.
Maybe there are specific FTTrN distributed power systems being developed in trials for BDUK (after all, VM do it with dedicated power cabinets and wires), but it's a nasty refit. RP makes sense.
Posted by andrew (Favicon staff member) about 1 year ago
NOTE the exchange and cab powered options mentioned, reverse power is still a little way off I believe.
Posted by Blackmamba about 1 year ago
Hi Broadband Watchers.
When these nodes ore fitted and who pays for then is it Openreach or a request from BT. If it is BT ISP will it be barred to other ISP,s.as they have paid for the installation.
I think it will be a joint build between the Goverment and BT thus giving access to all ISP,s is could be where the 7 year clawback money is used
Posted by andrew (Favicon staff member) about 1 year ago
@blackmamba Openreach is to deploy them on the same GEA, VULA arrangements as now, obviously different wholesale price.

So not a joint Government/BT exercise at all. Some BDUK projects may gap fund just as they are doing for FTTC of course, if it helps to meet the targets in a value for money manner.
Posted by Blackmamba about 1 year ago
Hi Euler
I would think the power would be from the (FTTC new cab ) that has a stand by
Battery fitted then via a D side to the node Postion.
This method was used over 30 years ago on line connectors using 50volts over the E and D side from the exchange this charge a battery in the line connector.
Posted by rtho782 about 1 year ago
The problem with RP is that the EU is obsessed with banning anything that uses power, and there are many people (e.g. my parents) that insist on turning routers off overnight (then complaining DLM has trashed their speeds).

So a 20 user node would be fine if all 20 left their modems on, and the EU didn't force the modems to turn off after 3 seconds of inactivity, but if everyone but you has their router off, and you are relatively far from the node, suddenly you're going to lose connection.
Posted by andrew (Favicon staff member) about 1 year ago
NOTE EU does not force modems to turn off, it mandates a low power mode which is a different issue.

If there is a EU doc saying turn off after 3 seconds of inactivity it would be an interesting read.
Posted by Blackmamba about 1 year ago
Hi Andrews staff
Thanks for reply this gives hope to 6k customers in Surrey that have been calculated to been under the (15 meg target at post code) at yesterdays date. These unit will be very cost effected in many location due to the fibre runs that are in position
Posted by TheEulerID about 1 year ago
@Blackmamba

Of course it's possible to provide power over phone lines, but it's very limited. 24 AWG copper wire measures about 17 ohms per 100m. As a loop doubles the copper length, then it's effectively 34 ohms per 100m, or 340 ohms per km. So as little as 50mA will have a 17v drop due to resistance in the wire (and deliver only 30V & 1.5W to the load.
As a VDSL2 line takes about 2W, this is a problem.
Posted by TheEulerID about 1 year ago
An addendum to the above. If the cable is similar to CW108 spec (0.5mm diameter), then the x-section area is roughly double and round-trip resistance is about 17 ohms/100 metres.

This means it's just about feasible to deliver 3W at 1km or 6W at 500m over a single pair (albeit it's at about 66% efficiency, so around 50% more power is required at the source end due to resistive losses).
Posted by cyberdoyle about 1 year ago
Gfast. AKA #gfarce. Another patchup, expensive, futile. And probably with public money. https://neil-fairbrother.squarespace.com/blog/2013/7/1/gfast-a-high-speed-cul-de-sac
Posted by Blackmamba about 1 year ago
Hi Euler
I think the Unit will be at a distance of a mile from the Cab which will require fibre so either double up the D sides or provide an extra power feed with the fibre which may go via ducting or overhead.
Posted by max360 about 1 year ago
I guess BT would only be enabling g.fast for the existing FTTC enabled cabs. And not for under review (UR) cabinets?
Posted by WWWombat about 1 year ago
Forward power is of course possible, and likely to be used on early trials; getting reverse power right is probably tricky - and the initial trial boxes will probably be more power hungry than the eventual commercial units.

@rtho782 The design aim is to make the node intelligently powered, so that just enough hardware works when only one user provides power. The G.fast protocol is also designed for discontinuous transmission, to aid power reduction when not in use.
Posted by mklinger about 1 year ago
BT should first make sure that all those that can be connected to superfast are connected, like many of the exchange only lines left behind.
Posted by Gadget about 1 year ago
@CD - the current announced plans are not with public money, and not as expensive as going to full FTTH, and unless someone else is prepared to stump up more money its another step towards FTTH so hardly futile.
Posted by fastman about 1 year ago
jacko -- if your can is not covered b BDUK you could always self fund it with your community
Posted by TheEulerID about 1 year ago
#CD
have you considered going beyond the playground level of insults and actually engage in real issues about the advantages and disadvanges of different approaches. You are sounding like a record with a stuck needle.
Everybody knows, that in an ideal world, fibre would be run everywhere. But in the one where we live finance and resources are limited. B4RN's volunteer supported approach is not replicable in most of the country.
Posted by otester about 1 year ago
Should cover most urban areas, if you live out in the country then you can pay for FoD.
You must be logged in to post comments. Click here to login.