back to article UK's BT starts trials of new hollow-core optical fibre networks

BT has started trials of hollow-core fibre optical networks to test if they can raise fixed-line speeds and produce mobile networks with lower latencies. The telco is performing the trials at its R&D campus in Suffolk's Adastral Park and aims to deploy up to 10km of fibre. It also intends to test the suitability of hollow-core …

  1. Eclectic Man Silver badge
    Boffin

    Scattering

    The article states "As light travels faster in air than it does when pushed through solid glass, latencies are reduced." But I suspect that there would also be a gain in a reduction of light scattering in a hollow core, rather than glass, however pure.

    Nice to see the national carrier investing in infrastructure.

    1. Tom 7

      Re: Scattering

      When I was there >40 years ago I was told the solid core diffuses into the guide making a lens that tends to encourage a more tight 'beam which results in less spreading of the pulse. The hollow core allows for longer paths and so spreads the pulses. Like I say this was over 40 years ago and we were working at 9.6Ghz so the physics may be a lot different using modern glass pulling and better lasers.

      1. Anonymous Coward
        Anonymous Coward

        Re: Scattering

        I think what they are using will be more modern; fibre design has changed quite a lot; it's gone all the way through photonic crystal fibres and out the other side...

        1. Steve K
          Coat

          Re: Scattering

          Sounds like they are using Polo-ised light

          1. Graham Dawson Silver badge

            Re: Scattering

            I'm sure there's a hole in this plan somewhere.

          2. Fruit and Nutcase Silver badge

            Re: Scattering

            they'll be mint-ing the potential rewards

        2. Trubbs

          Re: Scattering

          Sounds mint

          1. Will Godfrey Silver badge
            Happy

            Re: Scattering

            Robert might not agree

    2. Roland6 Silver badge

      Re: Scattering

      If latency really is that important, why not go the whole way and ensure the hollow-core is a reasonable vacuum...

      1. Ken Hagan Gold badge

        Re: Scattering

        A fair question, but I'm guessing that the answer is ...

        Because the refractive index of air is one point oh fuck all, so you are already enjoying 99% of the possible benefits and getting a long-lived vacuum seal on a thin tube several million times longer than it is wide is Quite Hard.

        Also, if memory serves, there is something called Knudsen flow which in this case means that evacuating the tube would be quite hard too. But my memory is definitely flaky on this point so I expect that someone who can be bothered to google will shortly be putting me right.

        1. Roland6 Silver badge

          Re: Scattering

          My query arose more from two considerations. Firstly, considering the production of hollow-core fibre - something I remember seeing demonstrated in the late 1980s, where the filament production process would naturally lead to a hollow-core with potentially near-zero air content.

          I guess a reason to not advertise it as actually having a vacuum, is that given the typical amount of splicing and termination, maintaining the vacuum will be expensive and problematic, although Knudsen diffusion might assist in long lengths. So performance could be better than light passing through the air.

          The second consideration was a little frivolous, namely one of the user groups that will be interested in

          sub-millisecond speed improvements would be financial traders. So whilst they would appreciate the big improvement over solid glass fibre, they would still see the window that might allow someone else to beat them to a trade...

          1. Ken Hagan Gold badge

            Re: Scattering

            Ah yes, the high-speed traders. Probably the only people who care and equally probably able to afford it on the back of their ill-gotten gains.

            1. Eclectic Man Silver badge

              Re: Scattering

              Just a thought: Re vacuum, one of the ways to tell when a crystal is glass rather than a precious or semi-precious stone is that glass contains small or tiny bubbles. In the event that the hollow core is a high vacuum, eventually the bubbles in the glass could rupture, which might reduce the efficiency of the fibre with debris in the core. Admittedly I have no expertise in this area of knowledge, so perhaps some physicists could help out?

      2. Fruit and Nutcase Silver badge
        Alert

        Re: Scattering

        Can you get a sheep that is small enough to check on the speed of light inside the hollow core?

        Velocity - (Percentage of maximum velocity of sheep in a vacuum)

  2. katrinab Silver badge
    Paris Hilton

    "Previous tests from the University of Southampton showed a reduction of 30 per cent, although BT said it believed they could drop further, up to 50 per cent."

    I thought the speed of light in fibre was about 0.65c to 0.7c? So while it is theoretically possible to get a 50% increase in speed, I think it would be unlikely, because while the light will travel faster, I don't think you are going to be able to get it to travel in a straight line through the pipe. It is still going to bounce off the walls, so actual distance travelled will be longer than the length of the cable.

    1. steven_t

      Latency not speed

      The article says BT believes it could achieve a 50 per cent reduction in latency, which isn't the same as a 50% increase in speed. Unless I've misunderstood how they think it is possible, it would require a 100% increase in speed.

      The speed of light in a vacuum is pretty fast but it comes a distant second to the speed of light in a marketing pitch.

    2. Ken Hagan Gold badge

      For a sufficiently thin fibre (comparable to the wavelength of the light) I think it does travel through in a straight line. You have to start thinking about waveguides and solving Maxwell's equations for the whole interior volume of the fibre.

      1. SCP

        It is very interesting stuff (I had a quick look at it last year as part of some other work). Congratulations to the Lumenisity and UoS teams for getting this technology developed and into trials. Hope it goes well.

    3. Jellied Eel Silver badge

      I think it depends how you read it. So I tend to work on traditional fibre being 0.7c or around 4.7ms/1000km. So I'm assuming the 30 & 50% numbers are for the remaining 0.3c, so around 0.85c. It's funky stuff, and I remember when Soton first announced it, but production runs were in the cms. They were also doing some interesting multi-holed cores, which I guess wouldn't pass a dope test..

  3. zuckzuckgo Silver badge

    Frickin' lasers

    > "and an ability to withstand higher power levels, making them well-suited for use with lasers."

    I designed fibre optic transmitters in the early 80s and even then we only used laser diode emitters for telco equipment. Have they been using anything else for long haul since then?

    At the time, T1 (1.5 Mb/s) was commercially available with T3 (45Mb/s) as the goal for new telco products. By the time I moved on we were building specialized 50 Mb/s local fibre networks (Hubnet) and trying and compete with Token ring and 10Mb/s Ethernet.

    1. the spectacularly refined chap

      Re: Frickin' lasers

      T1 was twisted pair copper...

      1. Roland6 Silver badge

        Re: Frickin' lasers

        >T1 was twisted pair copper...

        In the 80's T1 over fibre was created and was used to replace copper in the telco backbone.

        BT similarly began the upgrade of its core copper network to fibre...

    2. tip pc Silver badge

      Re: Frickin' lasers

      E1 over here, effectively bonded 64kb ISDN, 24 for T1 in the US & 32 for Europe providing 2mbps,

  4. Anonymous Coward
    Anonymous Coward

    Cool...

    So another amazing technology coming our way that we'll access through a shitty Intel Puma powered doorstop.

    ...and after the bean counters get involved it'll be all bog roll tubes and maglites to save money.

    1. tip pc Silver badge

      Re: Cool...

      All over wifi

  5. Ochib

    High-frequency trading

    If there is an advantage to hollow fibers HFT companies will be the first to move

    1. gotes

      Re: High-frequency trading

      Hollow fibre trading?

      1. katrinab Silver badge

        Re: High-frequency trading

        High frequency trading. Those extra nanoseconds can make a huge difference to their profits.

        Gaming is pretty much the only other use-case where shaving off milliseconds from latency helps.

    2. noisy_typist

      Re: High-frequency trading

      Only where it can beat microwave.

      A transatlantic cable would be interesting if someone hasn't already done it.

  6. Anonymous Coward
    Anonymous Coward

    Good stuff but Adastral - really? Having visited the place several times, the teams there, whilst good, don't see the full operational practicalities. Hollow fibre - how will that manage with line breaks or damage? Assume the core of the cable around the hollow fibre is strengthened to avoid damage.

    Meanwhile, the OpenReach team can't even update my cabinet in the centre of town, meaning no one in the street can get fast Internet (despite speed promises, it was only ~5Mbit/s). We have the choice of Starlink or Virgin Media. Both are much more expensive.

    All we get is this;

    We're sorry, our fibre products aren't available for you yet.

    We don’t have plans to upgrade your area yet to our fibre services, but you may still be able to get our Standard copper products - check with your provider of choice.

    We're upgrading over 20 million homes and businesses over the next decade to our Ultrafast Full Fibre network, giving speeds of up to 1Gbps.

    We update our build plans every three months, so provide your details and we'll keep you up to date if anything changes.

    Also, why not consider our Community Fibre Partnership programme to see if you could get Full Fibre sooner?

    :-(

    1. Anonymous Coward
      Anonymous Coward

      "Meanwhile, the OpenReach team can't even update my cabinet in the centre of town"

      That's a financial, rather than technological decision.

      1. Anonymous Coward
        Anonymous Coward

        It's not always financial - things like cobblestone streets and other historical features can make areas harder to upgrade than they may appear at first site or cause them to go through significant planning steps before they can carry out work. It doesn't stop it happening but can lead to significant delays.

      2. Anonymous Coward
        Anonymous Coward

        Agreed and they weigh up the value per street. It's entirely possible the next street is just fine but on your street there isn't enough demand.

        Virgin Media tend to send letters round offering to deploy Virgin Fibre to the street and offer a discount for it, but they require a minimum number of people to switch. My mother in law's street in Sutton had this letter and absolutely nobody (including her) signed up. They all moan about shit broadband (this particular street maxes out at 5mb/s) but they won't pay for it.

        It's got to a point now where broadband availability can have an effect on the value of your house, so it makes no sense to cheap out on broadband.

        Take North Ascot for example. There's some stunning houses there for sale, but they just won't sell because the comms round there is so fucking bad. There's a house there that I viewed 3 years ago that is still on the market. They keep taking it off and putting it back on, but it just won't sell...I declined to put an offer on because after I researched the Comms availability, I decided it it wasn't feasible to live there given my profession.

        1. Ochib

          I used to work in a village just outside Birmingham. BT said that they would run FTTC if there was more than 100 people would express an interest in that product.

          It took us 2 days and the local phone book to get 200 people signed up.

    2. Anonymous Coward
      Anonymous Coward

      How many users are connected to that cabinet, and how many have 'expressed an interest' in faster broadband?

      I moved house a few years back, and dropped from ~40Mbps to less than 20Mbps as the cabinet was still an old one. It was a new build house, but they connected them to an existing underutilised cabinet.

      Similar to you, no plans by OR to upgrade the cabinet at the time.

      From what I found out (various discussion forums etc), basically OR prioritise the ones they think will pay them back quickest for the investment, i.e. the higher the number of connected lines, especially if they've shown interest in faster broadband, the higher up the priority list. Once high enough, the upgrade gets planned in.

      My guess is that they have some threshold, i.e. does cabinet X have > N people interested in faster broadband, if yes, add to the upgrade list. With the threshold gradually being reduced each year as more cabinets are upgraded.

      Locally, we basically asked everyone in the area to register their interest in faster broadband, even if they weren't actually bothered about it.

      No idea if this had an impact, but around 10 months later, our cabinet ended up on the job list, and got the actual upgrade to FTTC about 6 months after that.

      1. Anonymous Coward
        Anonymous Coward

        "From what I found out (various discussion forums etc), basically OR prioritise the ones they think will pay them back quickest for the investment, i.e. the higher the number of connected lines, especially if they've shown interest in faster broadband, the higher up the priority list. Once high enough, the upgrade gets planned in."

        "payback" due to a higher number of connections is also the equivalent of "providing more services to more end users faster" in a large scale rollout.

  7. Panicnow

    Competition from Starnet et al

    Having watched and competed with BT. Their policy is to sweat legacy systems as long a possible, while giving lip service to government policy.

    They are only ramping up FTTx because they are worried about competitors.

    e.g. The LEO stuff, could be faster than traditional terrestrial, taking the premium market, so beloved of BT.

    That they wave the Union Jack to get tax payer money to do what any normal business would have to get investors for is particularly galling, to us innovators.

    Hey Ho, Nothing changes in big business and politics

  8. Anonymous Coward
    Thumb Down

    Every announcement and advert from BT...

    ...is about as trustworthy, honest, accurate and sane as every political party manifesto.

  9. rg287

    An edge over altnets?

    Although hollow-core fibre has yet to see any extensive real-world deployments, its use might provide BT with an edge in the face of newfound rivalry from alternative fibre networks (commonly known as altnets).

    I hardly think so. Altnets like HyperOptic, CityFibre and B4RN are either successful because they provide >1Mb service to places OpenReach aren't/won't go, or because they restrict themselves to multi-tenant buildings which already have a distribution point in the basement, making rollout cheaper.

    Whether your connection is via solid or hollow core is entirely irrelevant for the 99.99th percentile of consumers and businesses. Hypothetically, if the premises are a full 5km from the nearest active equipment, swapping conventional fibre (at 0.7c) for a magic product which manages a full 1c would save you all of 4μs of latency. Nobody is picking BT vs. altnet based on that - they'll just pick the cheapest provider at their chosen speed point (and then fail to get decent performance anyway because their wireless is horribly congested).

    Neat tech though and very interesting for the high frequency trading crowd, along with anyone in the business of long-distance backhaul. Chasing latency improvements becomes very worthwhile on a 4,000km trans-ocean link, and lower attenuation plus a tolerance of higher input power could reduce the number of mid-cable repeater segments required, lowering costs on submarine cables.

  10. mark4155

    Hollow as Openrech [sic] does best...

    Yes it's me again, Mark from Manchester on behalf of those who live in close proximity to city centres who don't have even fibre to the cabinet let alone to the premises.

    Soke to an OR engineer yesterday who confirmed my cab (street cabinet) did not have fibre connected. He reckoned OR had "missed it out", not his fault.

    So before you stuck up incompetent fuckers at Openreach start banging on about "Hollow" fibre try stop making hollow promises giving 2nd class excuses and third rate service.

    Here's my postcode M40 7AZ. Come round and sort your shoddy inept network out. I won't hold my breath.

    1. EnviableOne

      Re: Hollow as Openrech [sic] does best...

      government Huawei bashing hasn't helped either.

      the FTTC cabs that actually deliver have their MUXes in the other provider's stuff is awful.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like