Aug 28, 2014

Govt modelling says NBN is just fine — as long as you don’t want to use the internet much

The government's estimations of the download speeds we'll need in the future are laughable -- and alarming.

Stilgherrian — Technology writer and broadcaster


Technology writer and broadcaster

What, exactly, is the purpose of the National Broadband Network in relation to the following words? Enable. Suffice. Support. Demand. Need. Desire. Encourage. Surpass. Lead. Follow. Enhance. Boost. Revolutionise. Now keep those words in mind as you read the report "Domestic bandwidth requirements in Australia: A forecast for the period 2013-2023" from researchers at Communications Chambers, the modelling that underpins the government's NBN cost-benefit analysis. The modelling reckons that a decade from now, the median Australian household will be demanding internet download speeds of a mere 15 Mbps. Only households in the top 5% of internet usage will demand "43 Mbps or more". That's achievable with the latest copper-based technologies -- provided everything works as advertised, and the "or more" bit isn't too more-ish. And that means they'll all be happy with Malcolm Turnbull's Internet of Every Technology that's Not Fibre to the Premises because that's Labor and Wrong (IoETtNFttPbtLaW). "Mission Accomplished", right? Perhaps not. Communications Chambers has used a sophisticated-looking model, using 16 household types with varying numbers of adults and children. Three types of TV usage (SD, HD or 4K). Four levels of individual internet usage. That's 192 different household types all up. They've looked at the applications everyone in those households might use, how often, and how much internet bandwidth they need. They've tried to take into account the way different types of usage would interfere with each other, to give a peak internet speed requirement for each household. And finally they've looked at how that usage would grow over time. The resulting projections are substantially lower than others we've seen:
"A 2023 median demand of 15 Mbps may seem low, but needs to be seen in the context of the continuing benefits of video compression, and the fact that 58% of households only contain one or two people. Consider two people both surfing, both watching their own HD TV stream while each having a video call. Even this rather aggressive (and rare) use case only requires just over 14 Mbps in 2023."
As The Register noted, that's a rise of less than 50% over 10 years, a compound annualised growth rate (CAGR) of just 4.14%. By comparison, Cisco's Visual Network Index predicts a CAGR for 2013 to 2018 of 30%. Now every model is only as good as its assumptions. You could argue with Communications Chambers' assumption that each household uses half its total internet consumption in some random four-hour "busy period", or that children use half the bandwidth of adults, or that low-/medium-/high-usage households exist in the ratio 40/40/20, or that video compression will get better at a certain rate, or whatever. At least they've stated their assumptions -- although we don't have the computer model to play with, to see how changing those assumptions affects the result. But the key problem is the overall assumption that we'll see a gentle, incremental growth in internet demand -- whatever its rate for individual application -- based on the kinds of things we're doing on the internet today. During a digital revolution, they seem to have missed the revolution part. As just one example, the model completely misses the Internet of Things (IoT) , which will add billions of new devices to the internet in coming years, as everything that can have smarts built into it will do -- from airconditioners and light bulbs to toys and medical sensors. Indeed, the report almost seems to deny its existence as a concept: "The number of internet-capable devices may carry on rising, but as a practical matter a person is only going to be able to use a certain number simultaneously." Now many, perhaps most, IoT devices will use tiny amounts of bandwidth, but it all adds up. It seems perverse at best to leave it out. What else has been forgotten in this model? This brings me back to those words in the opening paragraph. If the NBN is built merely to cope with an incremental growth on what Australia as a society does today, then doesn't that mean it'll be incapable of supporting any revolutionary change that a massive increase in household internet bandwidth might bring? Previously I've written that the NBN is too bold for timid Australia. In response to a recent speech by Australia's chief scientist, I reckoned that Australian science and technology needs its own Team Australia.

Free Trial

You've hit members-only content.

Sign up for a FREE 21-day trial to keep reading and get the best of Crikey straight to your inbox

By starting a free trial, you agree to accept Crikey’s terms and conditions


Leave a comment

25 thoughts on “Govt modelling says NBN is just fine — as long as you don’t want to use the internet much

  1. bruce harris

    Yes. The whole issue of growth of demand studiously ignored in the fatuous whiteboard model presented by Turnbull.
    He went down considerably in my estimation after that stunt.

  2. Tom Gwynn-JOnes

    A government report coming out in support of government policy. Who’d have thunk it…

  3. Trog Sorrenson

    Turnbull may be smart, and he may be the best the Coalition have got on climate change, but he was stitched up by Abbott on this Communications portfolio, and now he is trying to stitch up the nation with his spin on the NBN.

  4. Nici

    We’re planning to get my late father-in-law’s home-made black and white TV working again. It seems timely. He worked on a wireless telephone system in the 1960s that was never implemented. Not much has changed.

  5. Patrick

    Any estimate of future internet bandwidth requirements that does not look like an exponential curve is a joke.

  6. Bob's Uncle

    Turnbull is no fool – he knows that technology is cyclical.

    By 2023, all the cool kids will be back to using their 33.3kb modems to dial into their AOL accounts.

  7. wayne robinson

    I’m currently getting an average of 50 Mbps through cable broadband. The App I use to measure my download speed claims that Telstra has an average of over 300 Mbps (I wonder who’s getting the speed?). 50 Mbps is too slow (let alone 14 Mbps) even though I rarely stream video.

  8. Ulor Boxo

    Is it not the nature of conservative political parties to look backwards?

    If God had intended us to use fibre optics He would never have given us the telephone!

  9. bjb

    Had a read of the “Domestic Bandwidth Requirements in Australia” report and can say it’s a complete waste of time. They focus very much on download bandwidth, and even then they make some heroic assumptions about future video compression. Their stated 1.5Mbps upload speed required (referencing the requirements Skype recommend) for HD video calls, although theoretically available, will likely be out of reach for most using FTTN.

  10. Duncan Gilbey

    What a let-down that blow-hard Turnbull is.

Share this article with a friend

Just fill out the fields below and we'll send your friend a link to this article along with a message from you.

Your details

Your friend's details