Broadband Forum Begin Certifying Gigabit Speed G.fast at 212MHz
The Broadband Forum has today announced that the first five G.fast devices running at 212MHz have successfully completed their certification program, which tested the kit at speeds of up to 1.45Gbps and copper line distances of up to 400 metres from the distribution point (cabinet/remote node etc.).
At present most of the G.fast (ITU G.9700/9701) hardware that has been deployed in the wild by telecoms operators (e.g. Openreach in the UK) tends to only make use of up to 106MHz of spectrum frequency, which works best at up to around 200-300 metres from a distribution point (beyond that the signal degradation rapidly kills performance).
However the technology has always had the potential to harness 212MHz of spectrum and a few years ago this was formally adopted into the standard via an ITU-T amendment (no.3). Sadly adopting 212MHz isn’t so much about improving G.fast’s limited reach as it is about boosting speeds to faster than Gigabit levels of performance on the shortest of copper lines.
Advertisement
Back in 2017 Openreach tested a Fibre-to-the-distribution-point (FTTdp) based G.fast setup that hit 1.66Gbps aggregate at the distance of 40 metres using 212MHz (here). In theory this approach could deliver an aggregate speed of around 1Gbps when installed up to 150 metres from a property.
Geoff Burke, Chief Marketing Officer at Broadband Forum, said:
“Delivering higher performance and gigabit speeds over a wider range of deployment scenarios, 212 MHz Gfast holds huge potential for operators. Demonstrating interoperability through Broadband Forum’s certification program gives operators confidence that the technology can be delivered at scale, reducing test time and accelerating mass deployment.”
Lincoln Lavoie, Senior Engineer at UNH-IOL, added:
“Operators are well-versed in the benefits of Gfast which enables ultra-fast speeds to the underserved Multi Dwelling Unit (MDU) market and is a powerful complement to fiber network buildouts. The advent of 212 MHz Gfast takes these services over copper to another level, aligning the copper service offerings with those of full fiber and further simplifying the operations of broadband networks.
This new certification program will help operators deploy this new groundbreaking technology quickly and more cost-effectively.”
Apparently the first hardware to pass through the forum’s new certification program, which is supported by the University of New Hampshire InterOperability Laboratory (UNH-IOL), comes from manufacturers including ADTRAN, Intel, NetComm and Nokia. More kit is expected to be approved as the year progresses.
In the UK all of this is fast becoming irrelevant. The local market has somewhat altered course over the past two years to focus on Fibre-to-the-Premises (FTTP) style deployments, which are more expensive and slower to build but offer significant advantages in terms of potential future speeds, maintenance and reliability.
Openreach has similarly scaled back their own G.fast deployment plan from 10 million to c.5.7 million premises by the end of 2020 (done so they can focus on bringing FTTP to 3 million premises by the end of 2020) and take-up of the new hybrid fibre technology is still very low (here). Admittedly the roll-out hasn’t been going for very long and ISPs aren’t putting much effort into marketing it.
Advertisement
Suffice to say that it remains to be seen whether Openreach will ever adopt 212MHz into their network, particularly with the focus now shifting so dramatically toward “full fibre” deployments. We did ask and will update if a response arrives.
Mark is a professional technology writer, IT consultant and computer engineer from Dorset (England), he also founded ISPreview in 1999 and enjoys analysing the latest telecoms and broadband developments. Find me on X (Twitter), Mastodon, Facebook, BlueSky, Threads.net and Linkedin.
« ISP Hyperoptic Brings 1Gbps Broadband to 32 Y&Y Residential Blocks
Latest UK ISP News
- FTTP (6095)
- BT (3668)
- Politics (2750)
- Business (2455)
- Openreach (2425)
- Building Digital UK (2350)
- Mobile Broadband (2172)
- FTTC (2089)
- Statistics (1929)
- 4G (1834)
- Virgin Media (1788)
- Ofcom Regulation (1599)
- Fibre Optic (1476)
- Wireless Internet (1471)
- 5G (1430)
I can’t see the point in the UK better keep going with fttp – though as ever price matters if it was super cheap perhaps there migth be some desire in limited cases
It’s all part of the superfarce to extend the life of the obsolete copper for as long as possible. There are not many people who darn socks any more.
@CC – End users don’t care how it arrives.
It is indeed there to extend the life of copper as part of a mixed technology migration – see: https://www.swisscom.ch/en/about/company/portrait/network/fibre-optics-network-expansion-map.html
It’s a viable option on very small loops and a progression towards full fibre.
It is not a viable option when deployed as Openreach have chosen to. Swisscom are doing it right, FTTC -> FTTS -> FTTB -> FTTP – they use G.fast in FTTS onwards, this will shine in FTTB scenarios.
On this I totally agree with @CarlT that the Swiss model is the way to go.
I never really understood what OR were trying to do with the GFAST PCP project. Well I do – play politics to prevent BT being broken up.
The counterpoint argument, to Carl’s, to using GFast with FTTB is that MDU’s seem to be a favoured target for Alt Net FTTP. If FTTP as a model works why bother with FTTB?
All of this is really in the rear view mirror now. If OR had done an aggressive rollout of GFast to MDU’s and focused on 35 profile FTTC as the stepping stone things would have made sense and been quick cheap and fast.
As it is my suspicion is that Gfast will turn out to be the ugly relative of FTTC. I don’t think most ISP want to bother with it as they just see it as a bundle of trouble as it will so rarely achieve its promise.
FTTS and FTTB using G.fast would’ve made sense. It is frustrating to watch BTOR use it to prop up a coverage numbers game. It’s not a bad technology if used appropriately. VDSL 35b could have helped considerably more people.
@TheFacts: End users DO care how it arrives, when they are rural.
When you are rural 400 meters is about 3-4000 meters too far, and you cannot get speed that 95% of the UK can get.
In our case BT/Openreach installed FTTPoD equipment then a few years ago replaced around 1500 meters the cable to newer copper because that was a financially viable solution to give us the bare minimum speeds (2-4MBit).
When you already have copper installed throughout the building as Swisscom would the case for using G.fast from the basement of it is obvious.
If you’re building ‘greenfield’ fibre or Ethernet over UTP to each unit is an obvious choice. Augmenting the existing copper via FTTB G.fast is a quick win and supplies 500Mb right now with more on the horizon.
I agree with @CarlT
“When you already have copper installed throughout the building as Swisscom would the case for using G.fast from the basement of it is obvious.”
However it ceases to be obvious when others are happy to install FTTP into MDU’s.
If OR had widely deployed out FTTC and then GFast or even 35 profile to the basement they would have secured a large slice of the market for themselves.
They didn’t and so are in the pickle they are in now.
TBH I still think OR would be better off freezing GFast and deploying 35 profile FTTC as it would have a larger audience. And yes it can only be done where there is no GFast and where there is a Huwaei DSLAM. But that does give a big estate where a fast and cheap upgrade for the many can be actioned with zilch civils.
@craski
“FTTS and FTTB using G.fast would’ve made sense. It is frustrating to watch BTOR use it to prop up a coverage numbers game.”
May I correct that statement
FTTS and FTTB using G.fast would’ve made sense. It is frustrating to watch BTOR use it to fail to prop up a coverage numbers game.
There fixed it for you.
Joking apart the time and effort expended on Gfast as deployed could have been better spent doing other things.
The trouble is that what has been deployed is the rump of the plan that Gavin came up with to get OFCOM off its BT break-up pathway. So it makes no technical sense only sense though the political prism political.
I doubt this will ever be implemented by BT, presumably all G.Fast pods already deployed would need replacing, they also do not mentioned anything about vectoring and how many connections vectoring can support at that frequency with current processing capabilities of the equipment, which might rule it out for a while longer in the UK where its supplied from the PCP and needs to support more than a dozen or so simultaneous connections.
G.Fast need to be K.Fast, kill it quick and concentrate on FTTP.
“I doubt this will ever be implemented by BT, presumably all G.Fast pods already deployed would need replacing, they also do not mentioned anything about vectoring and how many connections vectoring can support at that frequency with current processing capabilities of the equipment, which might rule it out for a while longer in the UK where its supplied from the PCP and needs to support more than a dozen or so simultaneous connections.”
I’d be very surprised if the entire G.Fast DSLAM’s needed replaced for 212MHz.
Newer line cards might suffice.
In fairness FTTdp G.Fast would off offered good speeds to everyone but it’s Openreach choice of doing Cabinet G.Fast that made it useless for all but a few. The technology itself isn’t bad it’s the way it’s rolled out that hampered it. It’s like saying DCOSIS 3.1 is bad because it isn’t FTTP but that’s not true either as DCOSIS 3.1 has already proved itself capable off offering 1Gbps across Europe and America. Most people will never care about how the speed is delivered just that it is and that it doesn’t hamper them. G.Fast FTTdp would of even been fine for rural areas.
Gfast is no good anywhere, and especially in Rural areas Matthew. There is no fibre for miles, and no cabinets, so if you were going to run fibre all that way you may as well finish the job properly and not bother with gfast and phone lines at all. It has only a very short transmission length. Why dig fibre 10 miles and then put it into old tired copper for a mile or two when it won’t work properly?
Not saying it would work for all rural areas Chris but some villages for sure where they could mount DP on Telegraph Poles run the Fibre to there and then use existing copper to hook up rest. As we all know it’s the final connections that are the expensive points to do. Not saying it’s superior to FTTP but it would be a significant step up from FTTC for rural areas. We have seen some FTTrN in BDUK projects which have worked would be similar principle but using G.Fast.
@Matthew
Has FTTrN been used in BDUK? I can recall it being used in a trial but not used in any scale to deliver targets?
Craski, the FTTC all-in-one kit are remote nodes.
@CarlT
Thanks, I didnt realise that AIO cabinets were being considered as remote nodes but can see that terminology could be applied.
I was under impression that the remote nodes referred to, were small underground or on poles and not cabinets as such.
https://www.ispreview.co.uk/index.php/2015/02/bt-fttrn-superfast-broadband-tech-trial-goes-live-north-yorkshire.html
Its a shame the economics did not stack up on the small nodes / FTTdp, that was the focus of hope for a while but that has gone now too.
GFAST is a total waste of time – the coverage is very inconsistent, we live in a small estate, on the same (short) road out of the 20 house only about 6 can actually get it, the houses that can get is also dose not follow logic eg:
1,3,7,11,15 can get it
5,9,13 cant get it
Same on our street.
The pole outside my front door is directly linked to the PCP. Our line is 140m and we get 290/48.
However those unfortunate enough to live closer to the PCP can’t get it as their pole is fed by an overhead link from our pole and so the line length is almost doubled.
As Gfast relies on the old copper layouts it is also hamstrung by then due to its short range.
Don’t get me wrong where it works it works very well.
Just for too few people.
Which makes the economics of it weird. As my suspicion is that the number of premises where it reakisticallyvworks is lower than the headline due to the copper arrangements.
G.Fast has a bad name in the UK because of the current way it is being used by OR. However it is successfully being used in Australia where the DPU is targeted around 150 meters from most premises and in Oregan US for MDUs providing G.Fast212 1 Gig service. There is still an argument that higher speeds could be delivered to more of us quicker using Fibre To The Curb and Fibre To The Street options and then re-engineer to full FTTP later.
For those of us that will be languishing in the FTTC wasteland until 2030+ it is an option I would still like to be considered. Yes the last leg will still be an achilles heel but I assume it is the last leg where the key factor in both resource and timescales of FTTP rollouts (cabling and civils) is.
G.Fast as used by BT is long range G.Fast, it’s a completely different thing really to how it was designed to be used. Vectoring is also a big issue, as getting a handle on half a dozen lines or so from a small node close to properties is doable, but vectoring over that large range of frequencies over 32 or 64 cards from a PCP, where line lengths are also considerably different from the shortest to the longest, has been a big challenge for the equipment manufacturers who have basically had to redesign the silicon for BT.
Even if G.Fast is used close to properties, it’s speed is good, but it still isn’t a brilliant technology, there is only so many hacks you can apply to get audio telephone cable that didn’t even support good audio, to work with data. It wastes a lot of energy spewing RF into the air, needs larges amounts of processing power for vectoring, and the infrastructure itself is old, often with bad joints external and internal to the home or insulation breaking down after 30 or 40 years, and for many homes long due for replacement anyway.
Its time to give up with telephone cable for data, its the law of diminishing returns, you might as well just go the whole hog and install FTTP rather than try and bring G.Fast nodes closer. This might be different for other countries that have better telephone cable infrastructure to start with, but in the UK, our copper telephone infrastructure is a bit of a mess and very old. Some stretches out on poles like AM radio antennas, some under ground all the way in ducts maybe in collapsed ducts, or direct in ground, or half in open air, half underground. Manhole covers full of water with 30 or 40 year old joints sat in them. Just look at the tops of telegraph polls to see the variety and mess of joints and connections, then imagine that repeated underground.
Its time this infrastructure was replaced anyway, lets get it done.
The ANFP for the BT network (ND1602) has recently been updated with the new 212Mhz spectrum. I’m taking it as a good sign that BT is planning on implementing it.
https://niccstandards.org.uk/publications/nd1601-1699/