» ISP News » 

BT Testing Co-Existence of G.fast and VDSL2 in up to 17.6MHz of Spectrum

Wednesday, January 11th, 2017 (8:56 am) - Score 5,801

One of the problems with BT’s new 330Mbps capable hybrid-fibre G.fast technology (ITU G.9700/9701) is that it can’t harness all of its spectrum (106MHz) because it needs to avoid interfering with the existing VDSL2 (FTTC) based “fibre broadband” service. But if they could co-exist.. speed boost.

At present the dominant VDSL2 (up to 40-80Mbps) service uses up to 17.664MHz of spectrum, while G.fast keeps its distance and uses spectrum between 19 – 106MHz in Openreach’s (BT) network. In this setup the two don’t have to worry about causing excessive interference and can operate alongside each other within BT’s network, but G.fast is clearly missing out on some performance.

The other problem is that both technologies use two different forms of half / full duplexing (i.e. the transmission of data in two directions, either asymmetrically or symmetrically). On the one hand VDSL2 uses Frequency Division Duplex (FDD), while on the other G.fast harnesses Time Division Duplex (TDD). Suffice to say that getting the two to co-exist within the same spectrum would be incredibly difficult, but perhaps not impossible.

Last year we became aware via the NICC, which is a technical forum for the UK communications sector that develops interoperability standards for related public networks and services, that BT was conducting a “ground breaking” technical study into the options for overlapping VDSL2 and G.fast spectrum up to 17.664MHz (slides). In December we got a little more detail on this, but haven’t been able to report it until now due to the intense amount of pre-Christmas work.

Apparently the key remaining problem is crosstalk interference from G.fast into VDSL2 and to what extent this difficult but predictable non-stationary crosstalk could be mitigated. Some readers may recall that crosstalk on VDSL2 (FTTC) broadband lines occurs where lots of active copper lines would effectively create interference for one another, thus Vectoring was introduced and this works a bit like noise cancelling headphones (i.e. it coordinates the copper line signals in order to remove most of the interference).

Some of our readers may recall that Vectoring had a troubled birth on Openreach’s VDSL2 lines and a variety of problems meant that it has only ever been implemented in certain very specific / busy areas (i.e. where the problem is quite acute), which can help to return FTTC speeds to more normal levels (here). By comparison Vectoring has been built into G.fast by default and works much more smoothly.

However BT’s technical bods are aware that true co-existence like this probably cannot be achieved without some trade-off in VDSL2 performance, although this assumes that the operator would stick with a purely static approach to spectrum management. We understand that BT is also looking at either a semi-static or dynamic option, although the latter might be too complex to implement.

At present all of this is still subject to on-going work and no final decisions have yet been made (each approach will require a complexity vs. benefit analysis), although the impression we get is that co-existence may soon become a very real thing and this would then be followed by a related update to BT’s Access Network Frequency Plan (ANFP) and guidelines. The result could be a nice speed boost for G.fast and perhaps a slight improvement to coverage.

So far as we’re aware BT appear to be the only operator that is currently considering all of these different techniques / options together and if they can pull it off then it would be a technically very impressive feat, although much may depend upon the penalty it introduces to VDSL2 performance (nobody likes to lose broadband speed).

The group is being conducted by BT’s Head of Access Platform Innovation, Kevin Foster, who is also Chairman of NICC’s DSL Task Group.

Share with Twitter
Share with Linkedin
Share with Facebook
Share with Reddit
Share with Pinterest
Tags: , , ,
By Mark Jackson
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 Twitter, , Facebook and Linkedin.
Leave a Comment
26 Responses
  1. Phil says:

    “However BT’s technical bods are aware that true co-existence like this probably cannot be achieved without some trade-off in VDSL2 performance”

    So more cross talk and reduction in speeds for VDSL customers that will never see vectoring, I suppose that is one way to force people to upgrade to G.Fast that are close to the cabinet. Those customers on longer lines from the cabinet that are desperate for higher speeds, that might have seen a small improvement from less cross-talk as VDSL users migrated to G.Fast that until now avoided the same frequencies, can now look forward to their speeds going in the opposite direction with added noise from G.Fast that is going to be harder for VDSL modems to work around given the time division multiplexing it uses.

    All this so BT can advertise ever higher top speeds that fewer and fewer people can actually achieve.

    We should have skipped G.Fast completely and gone straight to FTTP, the costs involved to just increase speeds for those that already get good speeds so they now get 100-200 Meg more down if they are lucky, and another 10 Meg up, which in reality for most will run at considerably less speed anyway due to congestion and throttling elsewhere, just doesn’t warrant the expense.

    Why didn’t BT, who are now only installing G.Fast at the cabinet, so no investment in bringing fibre closer to homes, just not enable VDSL at faster profiles, as the achievable speeds would have been similar to G.Fast, but for a smaller cost and easier to maintain and support. This upgrade to VDSL hardware could then have implemented vectoring as well. Speeds achievable now on VDSL profile 30a is 200 Meg down, and 35b is 300 Meg. G.Fast makes no sense at all.

    1. Mark Jackson says:

      I’d wait to pass judgement until we actually see this approach being tested in the wild, then we can assess for ourselves where the reality sits.

      I doubt BT would adopt anything that imposed a serious performance detriment on VDSL2 as that would also impact the Government’s “superfast broadband” coverage target.

      All this work is about trying to avoid screwing up VDSL2, while still giving benefit to G.fast. So I suspect any performance loss with VDSL2 might be very small indeed. We’ll see.

    2. MikeW says:


      Your final question is well worth asking: Why not choose Vplus (profile 35b) instead of G.Fast?

      The answer comes not from considering what the relative benefits of these two technologies are. Instead, it comes from considering what the operator intends to do next.

      If the operator wants to send fibre deeper into the network over the next 10 years, but only slowly, and to continue hybrid copper technologies, then G.Fast is the solution to choose. If they are going to stop the hybrid approach at the current cabinet, and choose FTTB/FTTP thereafter, then Vplus is the one to go for.

      So far, Swisscom are going for G.Fast, and TDC (Denmark) are going for Vplus.

      If BT were intending to only ever upgrade the existing cabinets, then Vplus would be the obvious choice. That they are going for G.Fast is an indicator that, at some point, they are going to be putting nodes deeper in the network.

    3. Phil says:


      “If BT were intending to only ever upgrade the existing cabinets, then Vplus would be the obvious choice. That they are going for G.Fast is an indicator that, at some point, they are going to be putting nodes deeper in the network.”

      I agree, and originally G.Fast was all about bringing the fibre closer to homes in order to increase speed for more of us and a lot of people were excited to think they were not going to be left in the slow lane simply down to distance and a decades old system of unshielded thin copper wires (if really unlucky aluminium wire) that was only ever designed for audio, and not good audio at that. BT changed their minds when they deemed it too expensive, and instead are now going to make the fastest VDSL connections faster leaving a lot of people exactly as they were before.

      G.Fast wasn’t designed to be installed in a cabinet with an aggregation of dozens of connections coming together from widely differing distances, that wasn’t its use case, so I wonder how effective vectoring will really be and how well G.Fast speeds will hold up as subscribers grow in a cabinet. Then we need to factor in that G.Fast will be sat alongside the VDSL DSLAM with the potential for general electrical noise to be added to the mix and potentially direct crosstalk from using the same frequencies if that goes ahead, it just becomes a mess of trying to manage spectrum and patch firmware’s in CPEs and at the DSLAM to try and get vectoring to cope, a waste of time and money.

      VDSL was a good investment, it brought fibre closer to homes, and shortened the loop length considerably for many people, providing faster speeds to places that barely got anything before on ADSL, yet G.Fast doesn’t do that if all they are doing is installing it at existing VDSL cabinets.

      The deployment method should be the opposite of what they are doing, there should be legalisation in place that means it should only be allowed for BT to deploy G.Fast where it reduces line lengths and so brings fibre closer to homes and so paves the way for FTTP in the future for all.

      One day questions will be asked to why so much time and money was spent on G.Fast and why the UK is lagging so far behind and limping along still on corroded copper pairs. Yes they are a company that needs to make a profit, but are virtually a monopoly and are acting like it. If there was more competition they would have to be replacing the copper pairs with fibre or data grade COAX in order to compete, rather than going for small incremental upgrades that in the short term provides the highest returns and lowest risk for shareholders.

    4. AndyH says:

      @ Phil – You do realise that a large number of significant ISPs around the world are implementing G.fast?

    5. MikeW says:


      I’ll reiterate: My point is that BT’s plans are only *currently* to deploy from the cabinet. There is absolutely nothing stopping them from also putting fibre deeper into the network at any time and place that they choose. Next week, year, decade. At intervals of 400m, or 300m, or 200m or 100m.

      Who knows what they’ll be doing in 3-4 years time?

  2. GNewton says:

    Why not replacing VDSL with G.fast once tested and available in appropriate areas? Less crosstalk issues between VDSL and G.Fast.

    1. DTMark says:

      Because G.Fast ‘runs out’ earlier along the cable run and beyond a certain distance it is slower, go further along and it won’t work at all.

      Similar to the way in which ADSL can work better than ADSL2 on longer and poorer lines.

    2. Ignition says:

      Same reason they haven’t replaced ADSL with VDSL.

      DTMark – once you switch off VDSL the range of G.fast can be the same or better, depending which knobs you tweak. It runs out earlier because it can’t use the VDSL spectrum. No VDSL notching, good to go.

    3. Steve Jones says:


      The only reason it runs out earlier is that the ADSL and (maybe) VDSL frequencies aren’t available due to cross-talk issues (as ADSL comes from the exchange, it’s much weakened and would suffer badly from cabinet-based x-talk which would be at higher power – note that cabinets a long way from the exchange are permitted to use more of the ADSL spectrum).

      The reason why g.fast & VDSL might be able to do so frequency sharing is that they are both cabinet-based services at about the same place in the network, so there is no weak attenuated system to protect.

      A pure g.fast cabinet network would have (through appropriate spectrum planning) have at least as much reach as VDSL, maybe better given it has better vectoring. If, like LR-VDSL, it could fully use the ADSL spectrum, it would be better again.

      Of course, it’s not magic – there would still be big speed drops with distance, but g.fast is capable of working at long distances albeit at lower speeds with the right spectrum plan.

      The reason why there couldn’t be a wholesale switch to g.fast will largely be the disruption and cost of swapping out all the innards of the existing cabinets and replacing all the end user equipment. Maybe it will happen in time; who knows, but not for some time.

    4. MikeW says:

      G.Fast doesn’t (yet) have the range of VDSL2 – even if it didn’t have to coexist – because it is defined to be lower power.

      Even BT’s proposed increase of maximum power (from 4dBm to 8dBm, as part of amendment 2) still leaves it a long way short of the 14.5dBm allowed by VDSL2 cabinets.

      There are points where ADSL2 (not +) is better than both ADSL and ADSL2+ on long lines. The same (smaller) frequency set as ADSL, but different modulation, IIRC.

  3. Will says:

    And BT waste more time trying to solve the many issues that occur when you sweat copper lines! All this money and time spent on copper broadband tech which could’ve gone into FTTP investment which is undoubtably the more future-proof and long-term cost effective solution. The sooner Openreach are separated from BT the better. The sooner more of their business is lost to FTTP altnets the better.

    1. AndyH says:

      Why do we need FTTP now?

    2. Steve Jones says:

      The amount of money spent on investigating this will be in the low hundreds, or even tens of thousands of pounds. It’s a tiny fraction of the £20-25bn of a full scale FTTP rollout which would take a couple of decades and which no investor will fund, at least given the way the industry is regulated.

      OR will want results in a year or two, not decades.

    3. TheFacts says:

      @Will/GN – how will separating Openreach benefit FTTP rollout?

    4. GNewton says:

      @TheFacts: Who are you talking to? You keep asking the same lame questions on this forum. Yet you never have any sensible proposals on your own. I am afraid your fancy idea of having the government pay for a nationwide fibre rollout won’t cut it. There are good reasons for a legal or even a full structural separation of Openreach, you could read up on Ofom’s documentation at https://www.ofcom.org.uk/about-ofcom/latest/media/media-releases/2016/update-on-plans-to-reform-openreach

    5. GNewton says:

      @Steve Jones: The trials, including some limited field deployments, will cost more than what you suggest, though you are right, it is far less than a nationwide fibre deployment. However, there is no point in blaming the way the industry is regulated for the lack of BTs investment. BT has a large share in the mess the UK is in today as regards network infrastructures, Ofcom has clearly outlined why it thinks BT has not invested enough.

    6. AndyH says:

      @ Gnewton – Please explain why we need nationwide FTTP now.

  4. jeep says:

    Totally agree with Will on this, all the monies spent on the various technologies to prop up copper seems at face value to me to be a waste when the technology is available now pure fibre. I live approx. 350m from my cabs yet the max achievable speeds on my copper line(fed by pole) are 24m down & 7 up by the way which I am grateful for, so even g/fast etc would give me no benefit ? Am sure many other people will also have this problem ? .

    1. fastman says:

      The problems is the circa 400 + of the 550 + service providers only offer copper and openreach as part of its condition of licence is not allowed to force subscribers to move to fibre at detriment of a copper provider

  5. TheManStan says:

    I would hazard a guess that this would be appropriate for LR-VDSL and GFast cohabitation.

    With both the freqs of ADSL and the higher frequencies of VDSL2 relinquished, both GFast and VDSL2 can be improved in one fell swoop.

    However, that would require OFCOM to put an end date on ADSL…

    Going to a pure VULA system would them make ISPs think about PIA and their own infrastructure more… giving real competition to the market.

    It would also mean that BT would have to either look at more capacity for VDSL/GFast and increase the amount of FTTP, because the costs involved with so much extra capacity make FTTP more favourable.

  6. Adam says:

    @Phil Excellent Posts.
    Clearly put.

    BT have bet all their chips on G.fast. It’s a big mistake.

    Just because other Countries are rolling out G.fast does not make it a legitimate reason to do so in the UK.

    Outdated Firmware issues+rogue non-compliant devices, crosstalk aka “interference” (+low level industrial machinery “pump noise”), power issues are much bigger threats/issues than BT are letting on regards G.fast rollout and they are ignoring them, to save their own skin.

    But the big one, equivalent to Apple’s Bendgate (why it shouldn’t be the basis for any UK national infrastructure rollout) and what BT and other promoters of G.fast are failing to mention regarding G.fast is how susceptible this technology is to malicious interference/blackmail. Its just far too easy to take out G.fast Circuits.

    You have to ask why Ofcom has not even raised the issue to date of cybersecurity/robustness/safety of network for G.fast, regards outside malicious intent to disable it.

    Inteference which could be applied for brief short bursts, causing real disruption and difficult to pinpoint such sources, (either accidental or deliberate). Knocking off 100’s circuits at once. This could be done with cheap technology, placed at strategic places in the network, timed to operate at the same time. It’s a real threat and oddly, BT aren’t mentioning it.

    The security/reliability of banks of G.fast connections are very vunerable to blackmail in this way. Though not a known common threat at present, its certainly a reason to question BT’s thinking/approach in implementing G.fast Technology first and foremost for large scale UK rollout.

    BT’s so called ‘cheap approach’ (it isn’t if you want ubiquitous coverage) using legacy copper hybrid G.fast approach just isn’t a robust, safe one, you have to look at the potential threats in the World today.

    Personally, I think the malicious threat potential alone is enough to ditch this G.fast tech, going forward.

    G.fast is going to end up an expensive mess, fault finding problems mean install costs are outweighed by maintenance/malicious interference costs, over the longer term. It’s going to end up a real complex, tangled mess.

    Too many things are reliant on Broadband today. BT are being somewhat reckless “blinkered” in their approach building out the UK’s National Broadband infrastructure, where it can be easily attacked through malicious means.

    Attacking the G.fast network doesn’t even take direct access to BT infrastructure and the technology to do this is cheap. It’s biased technical thinking in favour of BT’s legacy copper network.

    It’s just not a sensible approach, its reckless.

    1. AndyH says:

      It’s the cheap approach? How much does it cost then?

      So what are your plans then? I guess 14 million premises with ultrafast by the end of 2020 (on top of VM and the altnet coverage) isn’t good enough for you…

    2. MikeW says:

      I’m interested in this hardware that can disrupt g.fast that isn’t already a threat to ADSL or DOCSIS. And works in such a widespread manner (while staying cheap) that it is better than, say, a pair of boltcutters or a JCB.

      Anyone who needs access (private, or internet) via an attack-free medium has plenty of other worries.

    3. Ignition says:

      So you could use devices that shoot out very broadband RF at a very high power level, having connected them to a 3G network and placed them next to a whole bunch of cabinets, or you could find out where the primary and DR sites are, get a couple of teams with bolt cutters, trace the manhole covers and cut the north and south or east and west high fibre count cables.

      If a business is relying on a single best-effort broadband connection for their connectivity with no fall-back they are asking for trouble regardless of the medium.

      This ignoring that DDoS carried out via the Internet and a ‘stress testing’ service is still considerably safer than physical sabotage.

      The kool kidz now are using SD-WANs and multiple, diverse broadband connections to augment, partially replace or even completely replace MPLS.

    4. 125us says:

      The same susceptibility to RF interference exists in all xDSL implementations too. It’s not a new problem and I’m not aware of a single instance where RF attacks have been implemented. As others have pointed out, far simpler just to attack the physical layer. I think finding any miscreants would be fairly easy, just by walking the cable route. AM radios with directional antennas are used today to find RF induced problems, caused by things like badly wired security lights and LED build attached to the wrong type of dinner switch.

      No-one who needs a high reliability service should really be using broadband anyway, the markets and technologies are different.

      I don’t understand why fault finding would be harder – is it something you’ve thought up, or is their real evidence of this from real world trials? It should be no more or less harder to fault then existing xDSL.

Comments are closed.

Comments RSS Feed

Javascript must be enabled to post (most browsers do this automatically)

Privacy Notice: Please note that news comments are anonymous, which means that we do NOT require you to enter any real personal details to post a message. By clicking to submit a post you agree to storing your comment content, display name, IP, email and / or website details in our database, for as long as the post remains live.

Only the submitted name and comment will be displayed in public, while the rest will be kept private (we will never share this outside of ISPreview, regardless of whether the data is real or fake). This comment system uses submitted IP, email and website address data to spot abuse and spammers. All data is transferred via an encrypted (https secure) session.

NOTE 1: Sometimes your comment might not appear immediately due to site cache (this is cleared every few hours) or it may be caught by automated moderation / anti-spam.

NOTE 2: Comments that break our rules, spam, troll or post via known fake IP/proxy servers may be blocked or removed.
Cheapest Ultrafast ISPs
  • Gigaclear £17.00
    Speed: 200Mbps, Unlimited
    Gift: None
  • Community Fibre £20.00
    Speed: 150Mbps, Unlimited
    Gift: None
  • Vodafone £25.00
    Speed: 100Mbps, Unlimited
    Gift: None
  • Hyperoptic £25.00
    Speed: 158Mbps, Unlimited
    Gift: Promo code: BIGBANG
  • Virgin Media £27.00
    Speed: 108Mbps, Unlimited
    Gift: None
Large Availability | View All
Cheapest Superfast ISPs
  • Hyperoptic £17.99
    Speed 33Mbps, Unlimited
    Gift: Promo code: BIGBANG
  • Shell Energy £20.99
    Speed 35Mbps, Unlimited
    Gift: None
  • NOW £22.00
    Speed 36Mbps, Unlimited
    Gift: None
  • Vodafone £22.00
    Speed 38Mbps, Unlimited
    Gift: None
  • Plusnet £22.99
    Speed 36Mbps, Unlimited
    Gift: £75 Reward Card
Large Availability | View All
The Top 20 Category Tags
  1. FTTP (4201)
  2. BT (3180)
  3. Politics (2148)
  4. Building Digital UK (2041)
  5. Openreach (1995)
  6. FTTC (1931)
  7. Business (1866)
  8. Mobile Broadband (1629)
  9. Statistics (1525)
  10. 4G (1397)
  11. FTTH (1372)
  12. Virgin Media (1301)
  13. Ofcom Regulation (1250)
  14. Fibre Optic (1246)
  15. Wireless Internet (1244)
  16. Vodafone (939)
  17. 5G (923)
  18. EE (920)
  19. TalkTalk (831)
  20. Sky Broadband (794)
Helpful ISP Guides and Tips

Copyright © 1999 to Present - ISPreview.co.uk - All Rights Reserved - Terms , Privacy and Cookie Policy , Links , Website Rules , Contact