» ISP News » 

ITU Gain Consent for New 40Gbps Ultrafast FTTH Broadband Standard

Posted Wednesday, March 2nd, 2016 (8:14 am) by Mark Jackson (Score 2,548)
fibre optic cables eclipse

The International Telecommunication Union (ITU) has gained first-stage approval (“consent“) for a new set of international Fibre-to-the-Home (FTTH/P) broadband standards, which can support both 40Gbps speeds (NG-PON2) and symmetrical 10Gbps speeds (XGS-PON).

Strictly speaking ISPs can already offer similar performance over their existing ultrafast fibre optic networks and indeed we’ve recently seen Singtel in Singapore promote a 10Gbps (Gigabits per second) service using their 10GPON (Gigabit Passive Optical Network) platform (here). Back in 2012 BT also demoed 10Gbps using a similar setup (here).

However the new standards go beyond those (see below for details) and it’s always helpful to ensure that everybody is building / using kit that is created to the same specification, which ensures good general compatibility. Both of the new standards were developed by the ITU-T Study Group 15.

10 Gigabit Symmetric Passive Optical Network (XGS-PON / ITU-T G.9807.1)

The previous XG-PON standard only ensured an asymmetric speed of 10Gbps download and 2.5Gbps upload, while a symmetric connection like the new XGS-PON can push 10Gbps in both directions and that’s a very significant improvement.

The new XGS-PON is said to reuse “existing PON standards to the maximum extent possible” and its physical layer follows the XG-PON (ITU-T G.987.2) and 10GE-PON (IEEE 802.3) standards. Meanwhile the design of the XGS-PON protocol layer is based on NG-PON2 (ITU-T G.989.3) and XG-PON (ITU-T G.987.3), and its ONU management and control mechanism is specified in ITU-T G.988.

XGS-PON also operates on the same optical distribution network (ODN) as XG-PON. The typical distance between the optical line terminal (OLT) and an optical network unit (ONU) is 20 km, and one OLT is capable of supporting up to 128 ONUs.

The XGS-PON wavelength plan provides for co-existence with G-PON, XG-PON and NG-PON2. An XGS-PON system is fully backward compatible with XG-PON ONUs, allowing the operation of both XGS-PON and XG-PON ONUs under a single XGS-PON OLT port,” said the ITU without twisting its tongue in the process.

40 Gigabit Passive Optical Network (NG-PON2 / ITU-T G.989.2 Amendment 1)

The big news today is of course the ITU’s new 40Gbps capable NG-PON2 standard, which mixes together ITU-T G.989.1, G.989.2, G.989.3 and is said to be the “first series of standards” that will provide download speeds beyond the current 10Gbps.

The main enhancements with NG-PON2 are provided by its use of multi-wavelength operation and ONU wavelength tunability in both transmitters and receivers (this reuses some of ITU-T G.988 and its wavelength management follows the guideline in ITU-T G.9802).

ITU Statement on NG-PON2

NG-PON2 is based on a multi-wavelength, point-to-multipoint architecture, and its primary solution is time and wavelength division multiplexed PON (TWDM-PON). A typical TWDM-PON consists of four to eight wavelengths in both directions, achieving a maximum rate of up to 80 Gbit/s in each direction. Each wavelength is capable of providing a subscriber with optical access up to a rate of 10 Gbit/s, and the upstream/downstream wavelength is also capable of operating at the lower rate of 2.5 Gbit/s. The typical distance between the optical line terminal (OLT) and an optical network unit (ONU) is 40 km, and one OLT is capable of supporting up to 256 ONUs.

Apparently major operators are already testing this new system with the “intention of deploying these systems in the near future.” Operators implementing NG-PON2 will be able to reuse existing optical distribution networks (ODNs) deployed for previous generations of PONs.

The wavelength plan of NG-PON2 provides for co-existence with G-PON (ITU-T G.984 series); XG-PON1 (ITU-T G.987 series); radiofrequency video overlay (ITU-T J.185; ITU-T J.186); and optical time-domain reflectometer (OTDR).

Other Information

Sadly the United Kingdom probably won’t be seeing 10Gbps FTTH/P broadband networks on a truly big national scale for another decade or two and even then it will take a lot of time and money to deploy, although many of the FTTH/P networks that have been built so far (around 400,000+ premises passed) could in theory be upgraded to support the new standards (not that you need or could even fully use such speeds).

It’s worth noting that the ITU are also investigating the possibility of 25Gbps per wavelength over PON and they’re doing so with the aim of enhancing the capacity to beyond 100Gbps in the future, which would bring FTTH/P services it up to the sort of standard that some core network links already deliver.

Separately there has also been a third development today (ITU-T G.709/Y.1331 – “Interfaces for the Optical Transport Network“), which saw the ITU-T grant first-stage approval of a revision to a key ITU-T standard underlying the Optical Transport Network (OTN) that will enable optical transport at rates that go higher than the current 100Gbps. This is of more use for major core network links rather than domestic FTTH/P.

The new standard extends OTN with a flexible n x 100G frame format (OTUCn), which can be used for line-side interfaces up to 25.6Tbps (Terabits per second) over the next 15-20 years. Yes.. Terabits! Apparently the initial n × 100G FlexO standard (ITU-T G.709.1) will be fully approved by the end of 2016 and standards for n × 200G and n/4 × 400G FlexO will be ready for when next-generation 200G or 400G client optical modules become available.

Suffice to say, there’s a lot of capacity left in them thar fibre optic cables.

Delicious
Add to Diigo
Add to Slashdot
Tags: , ,
Leave a Comment
39 Responses
  1. Steve Jones

    It’s interesting to see how much further PON technology is being developed. There is a major divergence between the PTP fundamentalists and the PON proponents. The PTP proponents are proclaiming it the only true way forward and PON is a crippled technology, and some sort of conspiracy by telecommunication companies.

    The reality is that (for existing infrastructure at least) available duct space is often limited as it’s full of copper cable and that will have to coexist for a very long time. It therefore follows that it will be necessary to use that duct space as efficiently as possible, and PON is the obvious way of doing it (at least as far as the splitters).

    Developments to increase throughput further will surely continue into the future. For those that complain of contended systems, all residential networks have contention somewhere, it’s just where does it happen and, secondly, is it at a level that matters?

    • FibreFred

      I think it will be a very very long time before we outgrow PON and have to go p2p

    • Steve Jones

      I tend to think that for residential and all but very intensive business users, the right answer is probably never.

    • Never say never Steve, the people who created IPv4 had a similar problem with predicting the future 🙂 . Poor comparison, I’ll grant.

    • As far as expected bandwidth demands go these were held back by a combination of no new ubiquitous application and increases in CPU power allowing for ever more compression of video.

      Worth remembering that broadcast TV can take a wavelength on the fibre independent of IP traffic, too, so as far as moving video from DTH satellite or cable goes only VOD needs to traverse the IP network.

      We’re likely nearing the end of increases of demands of 2d video in terms of resolution, there’s only so much detail the eye can perceive on a screen and I imagine 4k will be about it for most homes with 8k for the odd one.

      I guess holographic images are the only place left to go after that.

      It’s easy for the population of sites like this to forget that the average home is only consuming about a megabit a second at peak times. Increase that 50-fold you’re still not making GPON break a sweat.

      As far as the odd extremely heavy user goes, there isn’t even a need to physically split PON networks to deal with them – they can be placed on XGPON 1 or 2 with no change to the network in the field just the kit either side with others on the split joining them as and when required, alongside even increasing the size of splits, power budgets permitting, as heavy users are moved to XGPON and load on GPON is relieved. XGPON is specified out of the box for 128 premises splits.

    • Bob2002

      @Mark Jackson

      >”the people who created IPv4 had a similar problem with predicting the future”

      Not really, for the record –

      “IPv4 was a lab experiment that escaped into the wild. There was always an intent to follow the experiment with a production version, and IPv6 is that production version.”

      – Vint Cerf

    • Ah but you see Bob, that just means they were even worse at predicting what would happen in the future 🙂 .

    • wirelesspacman

      Yes, and they also totally screwed up the migration planning! 🙂

    • Steve Jones

      The other think about IPv4 is that people totally underestimated the ability of clever engineering to extend its life by using NAT. Granted, it has reached its limits in certain areas, but people were predicting the imminent exhaustion of addresses a decade or more ago.

      You could also say the same about copper pair transmission for that matter. It’s amazing what can be squeezed out with clever technology. If you’d talked to a telecoms engineer 30 years ago and suggested 500mbps over a few hundred metres of copper phone cable he’d have thought you mad. Then there’s the massive jumps that have been made with DOCSIS on coax.

      At least in the case of copper there are fundamental limits. The physical limits of the combination of increased attenuation per unit length with frequency are fixed. Add in the practical limits of power that can be used and floor noise levels, apply a little information theory, and you hit the theoretical limits.

      In the case of PON, we are absolutely nowhere near the physical limits of what can theoretically be achieved, it’s simply the technology (and cost) of the ONT equipment. Hence my comment, that I don’t see residential users outgrowing PON as there is still huge scope for expanding capacity before the ultimate physical carrying limits are met. It’s more about the cost of the ONTs and that, surely, will come down over time.

  2. GNewton

    As the article says this is mostly of no relevance to the UK because none of the major telecoms here have any plans for a widespread deployment of fibre, certainly not from BT which doesn’t even offer Fibre-on-demand anymore, or has to rely on taxpayer’s money in order to serve 1/3 of the country with copper VDSL.

    • Jonny

      Change the record.

    • Chris P

      @GNewton
      what speeds do you think should be available to all?
      how much are you prepared to spend for access to such speeds on a monthly basis?

    • Pete

      BT are fast becoming a laughing stock . All that money (£800 million from the government I believe) and manpower and they still can get us into the top 50 countries with FFTX. Google the list and you’ll be crying with laughter at the countries that have better fibre penetration then we do. PON ha! Not when the likes of BT refuse to move from the dark ages of copper phone lines. “but fibre is more expensive” absolute rubbish, I’ve worked in fibre installation and now I train others and one of the advantages of PON is the lower installation costs.

      I’m so happy I switch to virgin with a HFC system a couple of years ago, just had a free upgrade to 152Mbps while my best mate living in a relatively new housing estate is till waiting for his Infinity to be upgraded to the 70Mbps he was promised almost 5 years ago.

      BT…not unless you paid me.

    • New_Londoner

      @Pete
      I think you’re confusing FFTX [sic] (assume you meant FTTX) with FTTH. On FTTX charts you would indeed see the UK towards the top, and certainly ahead of many other leading economies.

      As for your move to Virgin, you omitted to mention upload speeds. Also, you appear to be lucky that you’re not affected by congestion, bearing in mind many other posters on here have reported speeds on such lines slowing down to under 10Mbps.

    • GNewton

      @Chris P: “what speeds do you think should be available to all?”

      Each household, and each business has their own needs when it comes to speeds and how much to pay for it. The issue is that all too often it’s not possible to order higher speed broadband, even when a customer is willing to pay for it, simply because it cannot be provided in the UK everywhere, it’s often like a postcode lottery.

      A re-introduction of a nationwide fibre-on-demand could be a quick solution, without unduly burdening taxpayer’s or telecoms. The fact that BT is unable to serve a large portion of the country on a commercially viable basis is made worse by its inflexible ‘one size fits all’ approach (such VDSL in rural areas). This in an inheritance from its old GPO-monopoly thinking with its lazy “Can’t Do” culture. The efforts by the ITU to create higher-speed FTTH standards are applaudable, but largely irrelevant to the UK situation.

    • FibreFred

      Oh GNewton you post the same comments on tbb but as JNeuhoff but pretend not to be the same person. Poor show

      I’ve read everything you have posted above so many times before as have others hence their replies have you anything meaningful to contribute?

      It’s a good article and something bt will take up no doubt but as with all your other comments you simply use it to run down bt

      🙁

    • GNewton

      @Chris P: To answer your other question: “how much are you prepared to spend for access to such speeds on a monthly basis?”

      Less than a leased line price, but certainly more than today’s FTTP. I might also add that Carl Thomas (Ignition) has posted some interesting suggestions on how to improve infrastructure competition in his TelcoTorment blog though Ofcom’s proposals to improve fibre coverage doesn’t go much beyond better PIA and stricter Openreach regulation. Personally I think Openreach’s broadband prices are too low, thus less incentive to invest in fibre.

      BTW.: Please ignore FibreFred, he’s stalking again today with his usual accusations. he also accuses me of fake reviews on Trustpilot, and has called me names.

    • Chris P

      The thing is GNewton, you can get a leased line of any speed you want anywhere in the uk you want provided you pay. Broadband is the much cheaper shared access alternative that won’t always have the fastest speeds or availability at the price points consumers are willing to pay. It’s a round peg in a square hole, some bits are just out of reach or have poor coverage at the commercial prices. you keep asking for the impossible and blaming bt for not delivering, when non of the others are delivering either. You will never see sky, VM or talk talk rushing to provide high speeds to the last 10% as its not profitable, you’ll also not see them rushing to fill openreaches ducts when there is already capacity they can lease.
      The only way to drive last mile competition is to force VM to LLU to others.

  3. Great, though I’m hoping the ITU and manufacturers get what’s coming out of the back of the ONTs and in turn the networks in our homes somewhat higher than 1Gb at reasonable prices sooner rather than later.

    We’re in a somewhat ridiculous situation where we’ve >1Gb broadband services in the world that are bottlenecked by the GigE ports on the back of the modem / ONT / gateway.

    Need to see 2.5Gb and 5Gb Ethernet become finalised and mainstream, and 10Gb drop in cost somewhat ideally.

    • Jonny

      You probably know about this but there’s a body to work on MGig

      http://www.nbaset.org

    • Indeed, thank you nonetheless. There’ve been a couple of consortiums that’ve set about the issue. Hopefully results will be with us soon. At the moment the only kit is some pre-standard enterprise hardware that’s pretty expensive.

      I don’t mind paying £15 per port for a 2.5Gb switch or even £25 per port for a 2.5Gb-10Gb switch but £75 per port for 10Gb is a bit much.

  4. Noel

    Nice to see so many BT people in the comment thread, as usual.

    Arguing with abused BT customers in comments sections must be much more fun than actually helping them.

    For the record, the following massive tech titans have built out much better broadband than ‘Digital Britain’:

    Kazhakstan
    Moldova
    Belarus
    Russia
    Lithuania
    Latvia

    I could go on, but it’s so depressingly pathetic to all but the ears of the ‘BT Lifers’ and their irrepressible ‘can’t do’ attitude.

    But fear not! By 2025, if you live less than 100m from your local cabinet (most people don’t), you might be able to get a really expensive half-‘ultrafast’ service at ‘up to 500 Mbps’ (speeds may vary, upload speeds will be at most 1/10th of download speeds).

    Unless your line sucks, and a lot do.

    In which case…er…BT will ask the government for more money whilst paying their shareholders whopping dividends.

    • TheFacts

      For each of those countries please state the regulatory, competitive and funding arrangements.

    • GNewton

      @Noel: While the countries listed by you may have a higher percentage in fibre broadband, there is proobably a bigger percentage of superfast (24Mbps+) in the UK, though you won’t see widespread fibre broadband in the UK for a long time to come, certainly not from BT.

    • GNewton

      @TheFacts: “For each of those countries please state the regulatory, competitive and funding arrangements.” You could easily find them through a simple Google search. Not sure why that is important too, you are already using a superfast commercial broadband service anyway.

      What do you think of the ITU efforts for a new ultrafast FTTH standards? Especially in view of the situation in the UK?

    • FibreFred

      The ITU is “International” What has the situation in the UK got to do with a new ITU standard?

      Should they shelve it until the UK has more widespread fttp?
      Ban the UK from adopting the standard unless it increases fttp coverage?

      No need to keep up the name change denials I was onto you as soon as jn became gn, so was everyone else

    • I keep being told I work for BT but I still don’t see any money from them 🙁

      I think I may have said this before regarding other countries, however I hope your emigration goes smoothly, Noel.

    • GNewton

      @Ignition: “I keep being told I work for BT but I still don’t see any money from them”

      And I keep being told of bashing BT, or doing fake reviews 🙂

      Some of the trolls here can’t have a rational discussion on this forum.

      What do you mean when you said: “I don’t mind paying £15 per port for a 2.5Gb switch or even £25 per port for a 2.5Gb-10Gb switch but £75 per port for 10Gb is a bit much.” ?

    • TheFacts

      @GN – how many BT FTTP customers? Plus agreement with developers will see more. Hence relevance.

    • Ignition

      The cheapest 10Gb non-SFP switch I can find has 8 ports and costs over £600.

      There are I am sure cheaper products that have a couple of 10Gb SFP ports but then the cost of the modules has to be added on.

      However in my case I have no driver to purchase them as nothing faster than 60Mb will be available to me for the foreseeable; I was just looking.

      I had looked previously when there had been a chance of higher speed broadband being delivered here.

  5. wirelesspacman

    My issue with PON is not the speeds available, though they will always tend to lag behind PTP, but the inability to properly “unbundle” the fibres.

    Copper LLU in the UK has been a major regulatory success story, but the rollout of PON by BT prevents the equivalent happening with FTTH (well, BT FTTH anyway). Instead, it locks in the need to use a Bitstream type of service from BT, with all that that implies.

    • GNewton

      That’s why Ofcom is proposing a better PIA to accomplish better infrastructure competition.

    • TheManStan

      If that is a real issue then using BT infrastructure is pointless and the ISPs should build out their own infrastructure.
      The reason being OFCOM is so reticent to remove the copper USO.
      With that in place the majority of the capacity of all the duct is taken for copper provision and BTOR has to retain capacity to meet future copper USO.
      Until OFCOM pulls it finger out there will not be adequate capacity for any PTP infrastructure.

    • MikeW

      I disagree that copper LLU has been a regulatory success story.

      It *has* been successful in bringing new players in, and bringing prices down – but that is all predicated on sitting on top of a mature infrastructure with sunk costs.

      Unfortunately, it actively gets in the way of upgrading that infrastructure to something new, by reducing the financial returns available. Even Ofcom describe it that way now.

      You’re probably right that a PON reduces the ability to unbundle dark fibre – but what incentive does BT have to deploy such dark fibre if it knows the regulator is likely to take that step in 10 or 20 years time? They aren’t obligated to deploy new stuff that makes Ofcom’s job easy.

      In any case, BT’s fibre spine infrastructure does look like it includes a proportion for point-to-point use. It’ll probably be available in the future, but perhaps only within the business-grade regulatory environment.

  6. Steve Jones

    Yes, LLU was definitely successful in the short term at reducing retail prices but it has also been a big disincentive to network investment. Also, the LLU operators are in danger of having a lot of stranded assets if there is a very large move to FTTx services. It’s unclear what payback period the LLU operators have been working on, but the capex was probably significant.

    In any event, the issue of cheap copper prices disincentivising network investment has been the subject of a lot of economic analysis.

    https://www.etno.eu/datas/publications/studies/plumreport-costing-dec2011.pdf

    • wirelesspacman

      Totally agree about the issue of cheap copper prices, it is something I have been banging on about for years now with my consulting work.

    • FibreFred

      And you can blame dido and her race for the bottom for that.

    • wirelesspacman

      Am not sure you can to be honest Fred. The race to the bottom has specifically excluded line rental.

    • Ignition

      Remember TalkTalk offering free broadband with their line rental and calls package?

      That is why line rental was excluded. They are at least partly responsible for both the race to the bottom for broadband pricing and the use of line rental as subsidy for it.

      Seems even Ofcom have tired of this and their ongoing FTTPR rollout now.

IMPORTANT: Javascript must be enabled to post (most browsers do this automatically). On mobile devices you may need to load the page in 'Desktop' mode to comment.


Comments RSS Feed

* Your comment might NOT appear immediately (the site cache re-syncs periodically) *
* Comments that break site rules, SPAM, TROLL or post via fake IP/anon proxy servers may be blocked *
Promotion
Cheapest Superfast ISPs
  • Plusnet £20.49 (*30.48)
    Up to 38Mbps (Unlimited (FUP))
    Gift: None
  • SSE £21.00 (*41.00)
    Up to 38Mbps (Unlimited (FUP))
    Gift: None
  • Hyperoptic £21.00 (*38.00)
    Up to 100Mbps (Unlimited)
    Gift: None
  • Origin Broadband £21.58 (*31.58)
    Up to 38Mbps (Unlimited)
    Gift: None
  • Sky Broadband £22.40 (*27.40)
    Up to 38Mbps (25GB)
    Gift: None
Prices inc. Line Rental | View All
Poll
* Javascript must be ON to vote *
The Top 20 Category Tags
  1. BT (1586)
  2. Broadband Delivery UK (1157)
  3. FTTC (1024)
  4. FTTP (950)
  5. Politics (784)
  6. Openreach (735)
  7. Business (673)
  8. Statistics (660)
  9. Fibre Optic (651)
  10. Mobile Broadband (585)
  11. Wireless Internet (526)
  12. Ofcom Regulation (493)
  13. 4G (474)
  14. Virgin Media (450)
  15. FTTH (398)
  16. Sky Broadband (377)
  17. TalkTalk (351)
  18. EE (300)
  19. Security (254)
  20. 3G (225)
New Forum Topics
Helpful ISP Guides and Tips
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
Promotion

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