NBN: Fibre to the world

NBN: Fibre to the world

Summary: In this feature, ZDNet explores how fibre deployments across the UK, New Zealand, Canada, and the United States are being achieved, at what cost, whether they have been successful, and how they compare to Australia's NBN.


The United Kingdom

  • Project: Fibre to the cabinet and fibre to the premises

  • Area: 242,910 km2

  • Population: 62 million

  • Premises to be passed: 19 million

  • Percentage: 66 percent

  • Cost: £2.5 billion (AU$3.6 billion)

  • Government/private/mix: Private

(Image: BT)

Openreach, the wholesale infrastructure arm of British telco giant BT, is spending £2.5 billion to roll out 3 million kilometres of fibre and 50,000 new cabinets for a fibre-to-the-cabinet and fibre-to-the-premises open-access regulated wholesale network that will be completed by mid-2014. It will service 19 million premises, around two thirds of the United Kingdom, and the network already has 1.4 million active services across 50 internet service providers (ISPs), including the big three: BT, TalkTalk, and Sky.

Speaking exclusively to ZDNet, Opeanreach's director of network investment, Mike Galvin, said that BT would not have been able to justify the cost of deploying fibre to the premises across the UK.

"Fibre to the cabinet is considerably cheaper. It varies from site to site, but in brownfields, it is typically four times cheaper, maybe even more," he said.

"We would not have a business case for doing that if we were doing fibre to the premises, simply because of the additional cost. Even as it is, it is a long-term investment with payback in the mid teens of years.

"Fibre to the cabinet has made the deployment possible, as a non-subsidised deployment."

Galvin stressed that fibre to the home is best for new housing sites, but said that a cabinet deployment at a maximum of 1km from each premises significantly cuts down on the cost of deployment, and makes it roll out much faster. The company is currently passing over 100,000 premises per week.

"Fibre to the cabinet has made the deployment possible, as a non-subsidised deployment." — Mike Galvin

"There's a very strong argument that says fibre to the cabinet is a minimum-disturbance technology, and if you're equipping a city like New York or London, you're not having to dig up masses of streets and do a lot of retro fit work," he said.

"It's something that can be overlayed on the current network, so it is a much more reliable and quicker deployment."

The power required for each cabinet is about half of that required for DSL services today, and each cabinet comes with a battery backup and an alarm.

Galvin said that the network has the potential for a lot more development, and while speeds are currently a maximum of around 80Mbps down and 20Mbps up, the company is testing out vectoring, which would see the speeds increase dramatically. Galvin said vectoring would likely be deployed into the network later in 2013.

Mike Galvin (Image: BT)

"We see vectoring as one of the cornerstone technologies going forward that will extend the capability and the life of the fibre-to-the-cabinet product," he said.

One of the biggest criticisms of a fibre-to-the-cabinet deployment is that it uses legacy copper lines to the premises that are often described as being too old or damaged to be usable for high-speed broadband. Galvin rejected this suggestion.

"That's not our experience. I can't comment on what the Australian local network is like, but a lot of our network goes back to the great expansion in telephony services in the '50s and '60s, and that network is perfectly good for carrying these signals," he said.

While copper has an asset life of 30 years for telecommunications companies, Galvin said it can continue to be used long after that.

"It doesn't mean that after 30 years and one day, it stops working. We've got copper in our network that goes back to the 1920s, and it is absolutely perfect," he said.

"Age is not necessarily a guide. It depends how well the network has been maintained, how well it has been looked after, and the quality of the original installations."

But not everyone has agreed with BT's approach.

Earlier this month, the chief techology officer of one UK ISP Timico, Trefor Davies, reportedly had problems with his own connection to the network achieving only 6Mbps speeds, and it was only after a technician came and examined his copper line, that it was determined that the line from the telegraph pole to his premise had deteoriated and needed to be replaced.

Last March, the company's former CTO Dr Peter Cochrane (PDF) said that FttC was "one of the biggest mistakes humanity has made".

"It ties a knot in the cable in terms of bandwidth and imposes huge unreliability risks," he said.

"The number one fault problem with copper is water ingress. Fibre does not care about water. The fault level in an optical network goes down very low. You can reduce manning, buildings, power consumption and everything."

Cochrane admitted that most of the cost in going to fibre-to-the-premise was in the civil construction of building the pits and ducts to each premise, and he suggested the UK government could fund some of the smaller construction players to boost competition and reduce prices for BT to go to the premise.

"It doesn't mean that after 30 years and one day, it stops working. We've got copper in our network that goes back to the 1920s, and it is absolutely perfect." — Mike Galvin

Customers who are on BT's network via a cabinet but want the full fibre connection can pay for the privilege. Openreach is currently trialling a user-pays system, but Galvin said that the connection cost doesn't come cheap, and users can expect to pay up to £3,500 for the installation of a full fibre connection that would offer download speeds of up to 330Mbps.

Prices with BT start at £29 pounds per month for a 38Mbps service with 40GB of monthly use up to £37 per month on a 76Mbps service with unlimited data. Sky offers a £34.50 per month plan for unlimited data use.

Galvin said that the key to the network's success is its speed of deployment.

"It's been the fastest deployment anywhere in the world. That comes partly with the technology we're using, but partly because we're running a production line," he said.

"We think that's the secret to success. Business wise, it's been a success, as well. For the commercial deployment, there is no government subsidy, we've done it in a regulated environment, we have an open network, and we've made a success of it," he said.

A second initiative to cover the remaining one third of premises not covered in Openreach's rollout is being subsidised by local councils as well as the UK government and the European Union. There are around 35 projects, and BT has won all of the bids so far.

Topic: NBN


Armed with a degree in Computer Science and a Masters in Journalism, Josh keeps a close eye on the telecommunications industry, the National Broadband Network, and all the goings on in government IT.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • USA

    Considering the USA is the land of the free market i always find it funny there is a duopoly and i have the choice of time warner cable or slower dsl from verizon....and thats it.

    .....so much for "the land of the free"
    • Re: Considering the USA is the land of the free market

      Few markets need careful regulation in order to remain free. The US has somewhat neglected that side of it, to its cost.
  • Vectoring

    This article good, but still there are facts more. I was conference recently, Alcalu said 0.5m vectoring shipped and commercial lines working. 11 customers.
    Hans van Zitvlak
  • Transmission monopoly.

    Vectoring REQUIRES a monopoly at the electrical level too. Fibre, even GPON, holds the longer term possibility of un-bundling.

    Over the short term, there is a requirement for a monopoly at the transmission level with the NBN, but the longer term path, I recall roadmapped somewhere, did indicate the expectation that wavelength and physical unbundling are likely.

    This means that NBNCo over the longer term should shift to a model very similar to the current ADSL market where providers can lease a line (or wavelength) and operate their own infrastructure.

    Vectoring cannot do that. You will have one transmission monopoly forever.
  • Node to Wireless

    The alternative not being discussed is FttN + Wireless with the option of FttP as needed and cost effective. "Time" magazine reported last year that during the Obama re-election campaign, the organisers made the discovery that - wait for it - less than 50% of those under the age of 30 actually had a fixed line number! They live - quote - in the "cellular shadows". We seem to be using a philosophy of the post WWII time - the mass development of suburbia and the "gorgeous 'ome" - to quote Dane Edna - and so on - rather than looking at just how people live today. If I am renting a flat or house, as many younger people do, will I sign a 2-year ISP deal for NBN based services? No!

    Now, in addition, let's suppose I have my new NBN FttH will I rewire the house with Cat 5 or 6 cabling to make use of its speed or just let cables dangle over the floor from my NBN cabinet to my - well - what iPad / Android tablet, TCP/IP HD-LCD-TV, etc.? Guess what? I'll use my WiFi hub!!! Hold on? The "last metre" is wireless and more likely so with the massive acceptance of smart phones and tablets.

    So what about - rather than have everyone in the street trying to administer their own 802.11(x) wireless network - why not high speed wireless to/from the node? (Yes - we know all about congestion, etc. but remember that NBN line is also shared under GPON).
    Wireless from the node could also be used with backhaul from the node via satellite (See KaComm !) Ok - but what about 1080i HD video - OK - but are all home users connecting to the NBN cabinet via a cable to their lounge room or wherever? I doubt it! Once again - wireless.

    Let's rethink what makes sense - economically and with a faster delivery schedule (latest results from Mr Quigley seem to indicate that for many of us in outer regional areas, such as the Gold Coast with those old RIM boxes with their Telstra ADSL-2 "sort-of" "Top Hats", will be well and truly gone by the time we ever see a fibre cable from NBN.) No - we are NOT in the NBN satellite area.

    Time for a massive rethink!
    • Interesting points

      You make some interesting points, but to counter that, there are a few other things that need to be taken into consideration.

      Firstly, that 50% under 30 report. Namely that it conveniently doesnt go into any details about how many are still at home, how many are sharing, be it with friends or Uni, and several other key bits of information.

      It IS important to understand that more and more of the emerging generation is going mobile, but to leave out details to prove a point is only helpful in the short term.

      People have to live... somewhere. While they are there, they are still going to be using the net on a day to day basis. Yes, more and more net use is through mobile devices, but its been shown several times that over 90% of data transmitted (either 92% or 96%, not certain) is done over fixed line connections.

      Mobile devices are the fast food of net usage. You check in with facebook, or check emails, or some other quick relatively quick task that doesnt take long. But for anything taking more than 5 minutes, you wait til you get home and do it there.

      Then theres entertainment. You cant stream a show or movie to a mobile device when out and about. At home, yes (we'll get back to that), but out travelling, no. Its just not possible and it gets very expensive.

      Why? Mobile broadband and home network wireless are two very different things. One relies of a limited amount of bandwidth that needs to be spread about all users in the area (several square kilometers), so a premium is placed on it. The other is a similar amount of bandwidth, but only over about a 100 square meter area. Your home network.

      There is a difference. To expect one to account for everyones needs is naive and self defeating.

      Your second paragraph is good. Its important, and its where a lot of use will be done. But that use of your wireless hub is NOT the same as a 4G tower out in the city.

      And because they are different, wireless from the node to the premise wont work. You're asking everyone to drive down a 1 lane road at the same time.

      What makes sense? Rolling out a system that wont meet our needs in 5 years, or roll out a system that meets our needs for the next 50 years?

      One last thing to consider. 4G wireless (and friends) convert your wireless signal to a fixed line signal to do most of the work. To realise the speeds people seem to think wireless can deliver, the fixed line has to be that fast as well. And right now, FttN barely meets the current generations needs. It wont meet the next generations needs.
    • Not same thing

      Inside your house you can use WiFi and ethernet.

      We have 100mpbs and yes to get the full bandwidth to any machine I took some of our wireless computers to gigabit ethernet.

      The remaining machines on WiFi are functioning much better than before - for instance the watching of iView is now reliable over our WiFi where it was always dodgy on 16mbps.

      If theere is more than one user in the home then the benefits of the shared speed are not the speed to a single computer.

      WiFi demontstrates the reason wireless internet can never handle dense population areas. The bandwidth per device/user is limited due to the total bandwidth being shared.

      So no - there is no need for a rethink - there is need for a more educated user base to avoid such misunderstandings of the basic math and physics of the problem.

      GPON Fibre has a high bandwidth and it is shared minimally. The signal is constrained to the fiber and does not interfere with adjoining signals (like copper) or any other signals in the area (like wireless).

      In your house you may be sharing your WiFi channel with the householders. Alternatively you could be sharing your wireless internet with several hundred households. Which would you think works better?
      • Also reason

        I should mention that I replaced the wifi only where the wifi signal was variable between 2 and 20 mbps due to the nature of the building.
    • Towers

      It's already known that the number of towers under that situation would have to roughly equal the number of cabinets for either GPON or VDSL.

      That's roughly 50,000 towers.
      Do you want a 30m tower a the end of every street?

      Given the number of communities already trying to block the NBN's wireless towers, do you really think this is even possible?
    • Look behind the argument

      Logic question: Flat sharers are prepared to sign a one year residential lease but not a two year ISP contract? What happens if RSP contracts are reduced to one year? Hopefully under the wholesale/retail NBN model switching RSPs will be streamlined and much easier for all concerned.

      It’s one thing for 50% of under 30s not having a fixed line for voice but what about for all other purposes? If your fixed line data speed is sub-optimal reluctance to have one could be because the offering is not attractive...but if the service were better might it not get more use?

      I have found a good way to deal with the ethernet cable on the floor issue is to take the cable through the ceiling. Neatly finished holes in share house ceilings could make a such a house an attractive proposition for landlords and their prospective tenants.
      • There are Monthly providers

        One is Club Telco
        Completely transportable, leasing a holiday home for a month, just switch locations.

        Abel Adamski
  • Fixed Wireless

    Is being rolled out to place's in Country Australia, I'm assuming that will be Firbre to the Nobe and then Wireless to the home.
    • no fttn

      that's fibre to the wireless tower, the tower covers large areas just like a cellphone tower
      • Thanks for clearing that up!

        Regard the Fttn issue, but regarding the Wireless, my understanding is it won't be like a Cell tower (or for us Aussies mobile tower) Rather it will be fixed Wireless, ie each connection will be a dedicated Wireless connection.
        • Tried to post a link to the

          Information on the NBN co's website the AGAIN the profanity filter kicked in when it was not needed.
        • Yep

          It will fixed wireless, it's the exact same technology that Optus and Telstra are rolling out for mobile wireless data; that is LTE, or 4G in marketing speak. The difference between mobile LTE and fixed is that you have a fixed antenna on your house which has a direct line of sight with the tower, that way they can give a guaranteed service.
  • Woops

    Should be "Fiber to the node"
  • LOL at 1km


    100m 100 Mbps 5%
    200m 65 Mbps 20%
    300m 45 Mbps 30%
    400m 42 Mbps 45%
    500m 38 Mbps 60%
    600m 35 Mbps 70%
    700m 32 Mbps 75%
    800m 28 Mbps 80%
    900m 25 Mbps 85%
    1000m 24 Mbps 90%
    1250m 17 Mbps 95%
    1500m 15 Mbps 98%
  • FttN Power supply.

    One thing that must be considered with FttN is that every single one of the possibly tens of thousands of cabinets will need power. I have seen figure (either on ZDNet or somewhere else) that state possibly three new powers stations would need to be built to supply the electricity. Not to mention the costs of getting the power to the box, wich depending on its location, could cost from a few thousand to tens of thousands of dollars per FttN box. (allowing for digging up and resurfacing, under road boring, connection to the grid, traffic control, plant and material costs, etc.) Sorry if someone else has mentioned this, but I only had time to skim through the previous coments all the comments.
    Why Knot
    • FTTN - Not Exactly Green is it?

      The power consumption of the FTTN cabinets alone should be ringing alarm bells for anyone who wants to protect our environment.