Deconstructing Australia’s wireless/NBN fetish

45

opinion There was a beautiful moment in Telstra CEO David Thodey’s speech this afternoon at the annual Charles Todd oration which perfectly encapsulated Australia’s enduring obsession with the idea that wireless broadband technologies could somehow make the need for NBN fibre obsolete.

A few moments earlier, the executive had been waxing lyrical about the glorious future that wireless technologies would offer, noting that there were currently about 24 million mobile connections in Australia (smartphones, mobile broadband dongles, tablets and so on), a number that was expected to increase ten-fold by 2020 and perhaps to “a billion” by 2030, as cars, fridges and every other appliance were hooked up as well.

“Who would have thought the amount of data on our mobile network would double every year?” asked Thodey. And then, getting a laugh out of the audience: “I’m trying to double the price on it, but it’s a little hard.”

Then the executive mouthed the dreaded words.

“People often ask me: David, what do you think about the NBN?” he said, as the room suddenly grew slightly more hushed and journalists all around me instantly started scrabbling for their pen and paper or iPads to take down every nuance of the Telstra CEO’s NBN response. “I say it’s great,” Thodey said with a relaxed grin. “We do not see a world of mobile versus fixed, we see it as integrated. It will be integrated with wireless, because wireless is a fundamentally enabling technology.”

It’s a refrain which virtually everyone involved in Australia’s telecommunications sector has been mouthing constantly for the past 18 months or so. NBN Co chief Mike Quigley, Optus chief executive Paul O’Sullivan, Communications Minister Stephen Conroy and even his shadow, Malcolm Turnbull, have all acknowledged the reality that future generations of Australians already do and will continue to need both fixed and wireless technologies to get by.

And yet, as the press focuses this week on revelations that not just Telstra, but also Optus has provisions in its contract with NBN Co restricting how it markets its wireless broadband compared with the fledgling NBN fibre, it seems a little hard to take the ‘complementary’ argument completely seriously. Why, if the two technologies will exist side by side, do all parties involved in the NBN process continue to make such a big deal out of it? Why do there need to be clauses in the Telstra and Optus NBN contracts dealing with the issue? Why can’t industry commentators stop pointing out the incredible growth of in wireless adoption, and the flatlining of fixed broadband? Why does this theme come up again and again in the NBN debate?

To answer this question, we need to go back a few weeks to an article I wrote trying to push the NBN debate to a different level.

In the article, I argued that the reason that the NBN was so contentious as a national issue was related to technical limitations in the kinds of fixed broadband technology we have available to us. On the one hand, we have the existing copper network, which in practice usually limits Australians to speeds of up to 16Mbps, with mediocre latency and constant faults.

On the other hand, I pointed out, we have fibre, which represents complete overkill for the sorts of applications which most Australians need broadband for today, offering speeds of up to 100Mbps, which will only rarely be fully utilised, and awesome latency which only a few applications will ever take advantage of.

In this light, governments are right now faced with a difficult policy choice – there is no middle ground between these two technologies – but the middle ground is what Australia probably requires, at least in the medium-term. Hence we have the NBN’s universal fibre: Complete overkill for today’s requirements, but probably a useful technology to have bedded down when 2030 rolls around and the computing paradigm is a thousand times more advanced than it is right now.

Today what I want to argue is that the reason wireless technologies are so contentious in the NBN debate is that they potentially represent a realistic middle ground scenario to resolve the problematic dichotomy between ineffective copper networks and overkill fibre.

Again, David Thodey described this situation well in his speech this afternoon, pointing out that Telstra’s Next G network – which remains the most capable network in Australia – currently offered speeds of up to 42Mbps down and 21Mbps up, with 84Mbps on the horizon as the telco rolls out the Long-Term Evolution standard in its network (the first LTE base stations soft-launched on Telstra’s network last week).

“It probably will rival many other fixed-line technologies,” Thodey said.

But he also clarified his speed comments. “42Mbps on HSPA+ … it doesn’t go that fast, you all know that, divide it by three and you’ll get the average speed,” the good-humoured Thodey added. And on LTE, he laughed: “If you’re the only one in the cell, you’ll get 84Mbps.”

It’s this uncertainty about the current and next generations of wireless technologies which makes them a permanent fixture of Australia’s National Broadband Network debate, and the wider debate about how consumers across the globe will get access to broadband services over the next several decades and beyond.

Over the past few years, Australians have come to rely on wireless technologies in so many aspects of their lives – from our iPhones and Android smartphones in our pockets with their rich media access, to the laptop USB dongles which keep us connected on the road, and now to the current generation of tablets which are speedily replacing paper in our workplaces and in our lives, we’re a nation which loves wireless. And we are not hesitant to either praise companies which get it right (Telstra, and sometimes Optus) and damn those who it wrong (hello, Vodafone).

But right now there is a general uncertainty in the Australian population about wireless technologies. Our smartphones sometimes lose signal, streaming video via mobile broadband usually requires annoying buffering, and it can be an absolute pain in the ass trying to sync your tablet to Dropbox via 3G, when all you want is that tiny PDF document to show your boss or the board.

Too often, it seems just when you most want wireless to work, it drops out or is too slow.

Then there is the mysterious development of the technology, which appears largely governed by shadowy committees of organisations such as the IEEE, which took an age to ratify the 802.11n standard which many equipment manufacturers were already shipping in their routers and laptops. LTE versus WiMax was a vigorous debate for a while, and even now that most of the telcos have standardized on LTE roadmaps, the ultimate end of those roadmaps is so less clear than that of fibre. Will wireless eventually get to gigabit? The CSIRO seems to think so, and LTE seems on an ever upward curve. But the truth is that we really just don’t know yet.

This uncertainty around the role of wireless and its potential to fit in between copper and fibre plays strongly to the long-held idea that governments should not make hard and fast decisions backing certain technologies – an idea which the NBN policy clearly abandons as an irrelevancy.

We have only to look back at promises made by politicians such as then-NSW Morris Iemma a few years ago to blanket central business districts with Wi-Fi to realise how futile it is for politicians to try and pick technology winners. With 3G mobile broadband now universal in every CBD in Australia, it would have been foolishness indeed for any state government to waste money on building Wi-Fi hotspots everywhere.

And yet the thing about telecommunications policy is that it currently does require the Federal Government to make technology choices. We do have a national network based on one technology – copper – and the Government must of necessity choose where it will place its funding and policy efforts to resolve some of our current roadblocks.

Personally, I believe most people actually use fixed and wireless broadband technologies for completely different things. They’re not so much complementary technologies as they are adjacent. My smartphone has little relevance to my ADSL2+ broadband connection, and I don’t really need it to. My loungeroom media centre has few links to my multiple Telstra Next G mobile connections, and that’s fine.

Because of this, I think the Government’s NBN policy is the right one from a purely technological point of view – although I don’t approve of its commercial model. And I think most people feel the same. But that won’t stop wireless rearing its head up continually over the next decade as the NBN debate continues on and on and on. It will do so because the role of wireless in our society – the end game for the technology – is not clear yet, and will likely not become clear for decades to come.

In the meantime, it remains a tantalising, luscious-looking, question mark.

45 COMMENTS

    • Also, Thodey said spectrum is a huge issue. ie. Need to get as much data off the wireless cell as possible as often as possible.

  1. Why you no like da WiFi hotspots? :-)

    If anything, we are going to need more WiFi hotspots and more (or any) femtocells to take the load off mobile cell towers. Even assuming a lot of that is happening we are going to need more cells in dense environments to shift the traffic from wireless to wired ASAP.

    • Picocells with 802.11u authentication from the carrier.

      Put them in densely populated areas, railway stations, street corners of the CBD, etc, and people will automatically authenticate and the data load is taken off the mobile network and onto a potentially much higher data only network.

  2. wireless needs an efficiently-designed and limited, task-specific network of fibre backhaul.

    wireless doesn’t need fibre being shoved up under every Grandma’s front gardenbed.

    • It has been estimated that to provide a comparable throughput and coverage area to the current NBN plan, but in a completely wireless fashion would require approximately 80,000 wireless towers.

      I’m sure we’ll all be happy to see one in your front garden bed should that eventuate.

      • that’s just Quigley’s nonsense rubbish. yes, he got bored one Sunday, wondered how many wireless towers it would require to deliver Labor’s mandate and started playing with his napkin. the problem with that is no one has ever suggested that wireless is a complete replacement for fixed. he just made it up. after watching that NBNco presentation at Macq Uni, i’m starting to wonder whether if most of the FUD originates from NBNco itself, rather than self-organising diehard fibre geeks in the community.

        the significance of “wireless” as a competitor to “fixed” is this:

        i/ the average consumer has a limited budget to spend on telco services, $X;

        ii/ if he/she is spending a bigger chunk of that $X on wireless, there will be less discretionary dollars available to be spent on “fixed” products;

        iii/ the diversion of spending dollars from fixed to wireless is absolutely crucial because it determines the revenues that the fixed sector can generate and, as a result, the underlying size of the capital base the fixed sector can support;

        iv/ positive ARPU alone is not enough to justify the investment case for any network, the size of that ARPU or the degree of cannibalisation from competing networks is absolutely crucial.

        a lot of restaurants and clothing shops close down even though they still attract some level of clientele and foot traffic. look at Borders – they have never been short of people visiting their shops, browsing books, visiting cafe, etc. yet, they still went bankrupt. telco is no different.

        • You are so full of shit.

          If they are building 2,000 towers to cover the 4% of the population that falls within the current planned wireless footprint, how many do you think will be required to do 97%?

          Use your brain.

          • nice trolling.

            did you even read my reply? i didn’t dispute it takes 80,000 or whatever towers to deliver 97% wireless only. i was saying no-one’s suggesting an all-wireless solution. Quigley’s attacking a strawman.

            carry on immortalising yourself on the internet as a public fool.

          • Yes you did. You said it was “just Quigley’s nonsense rubbish”.

            If it looks like a troll, barks like a troll, and smells like a troll, it’s almost certainly a troll.

            Nice try, troll.

          • Yes you did. You said it was “just Quigley’s nonsense rubbish”.

            If it looks like a troll, barks like a troll, and smells like a troll, it’s almost certainly a troll.

            Nice try, troll.

          • *Yes you did.*

            no, i didn’t.

            “just Quigley’s nonsense rubbish”.

            Quigley’s “nonsense rubbish” is suggesting that the Coalition’s plan is 100% wireless, which it isn’t.

            he’s attacking an elaborate strawman.

          • caught with my what? i haven’t backtracked at all. as usual, you can’t engage in a debate without deliberately misrepresenting what other people say. only to end up looking like a total dill.

            carry on making a fool of yourself. it’s all archived for posterity.

            G’nite :-)

          • caught with my what? i haven’t backtracked at all. as usual, you can’t engage in a debate without deliberately misrepresenting what other people say. only to end up looking like a total dill.

            carry on making a fool of yourself. it’s all archived for posterity.

            G’nite :-)

          • The only person that is trolling here is you Michael, as was stated in the business case, the value that Quigley came up with was providing 100mbit wireless to all of Australia using what was is now outdated wireless technology

            Coalition never had a policy to put wireless everywhere

          • The only person that is trolling here is you Michael, as was stated in the business case, the value that Quigley came up with was providing 100mbit wireless to all of Australia using what was is now outdated wireless technology

            Coalition never had a policy to put wireless everywhere

          • @deteego.

            The business case/NBNCo also say the NBN will “pay itself off by 2034” and claim the Tassie stage 1 was finished “on time and under budget”.

            Do you also agree with them there too?

          • That’s not very polite Micheal, but you might just get away with it (again) being a cough-cough ‘ valuable named contributor’ to the NBN debate.

          • That’s not very polite Micheal, but you might just get away with it (again) being a cough-cough ‘ valuable named contributor’ to the NBN debate.

    • The copper already runs under “every Grandma’s front gardenbed”. Frequently (not always) it’s in a 25mm plastic conduit. Anyone digging on an easement is required to “make good” when they leave. Why this obsession with digging gardens?

      And no, the fibre backhaul doesn’t have to be that “task-specific”. LTE’s preferred interconnect is Ethernet, the same interface as will be presented on the NBN fibre.

      • i didn’t mean “task-specific” in that technical sense. apologies for poor choice of expression.

        what i meant is that the whole point of the superior economics of wireless delivery of broadband (for light users) is that you minimise the capital cost on laying or accessing physical infrastructure. of course wireless needs fibre backhaul, but you minimise the footprint of the backhaul and rely on towers for the “last mile”. aside from mobility, that’s the whole point of wireless in terms of minimising capital outlay. case in point: it only cost Telstra $1bln to build a nationwide 3G network.

        the NBN is essentially FTTH, or “fibre backhaul” right up to everyone’s front door. none of that is necessary to deploy wireless. more importantly, it defeats the cost superiority of utilising a limited fibre footprint to deliver broadband. hence, wireless does NOT need the NBN.

        i presume there’s nothing in the anti-competitive NBN legislation that prevents laying new fibre for wireless towers or base stations. first of all, a lot of the fibre backhaul from the FSAs to POIs and elsewhere is dark fibre leased from Telstra. secondly, there’s no way in hell NBNco will lease any fibre under its control cheaply to wireless operators as this would potentially cannibalise their fixed broadband product. they won’t allow enterprising telcos a way of bypassing the CVC or any other charges by implementing wireless bypass solutions leveraging off NBN infrastructure.

        • I suspect this is the main reason why CVC costs are set so high.

          AVC costs is easily bypassed.. CVC not so much

          • i get your point.

            but ultimately it doesn’t really matter whether it’s AVC, CVC, ABC or XYZ…. bottomline, laying fibre all over the country is so ridiculously expensive, unless the Fed Govt writes off the capital investment, NBNco will have to levy some structure of fees to recoup the investment and the access regime will have to be watertight such that the aggregate amount of revenue required to service the debt isn’t bypassed in any form whatsoever.

            the SAU even positions NBNco to raise the nominal cost of accessing premium services over time by as much as 5% above CPI on an annual basis.

  3. “currently offered speeds of up to 42Mbps down and 21Mbps up, with 84Mbps on the horizon”
    Why are there clauses to say Telstra and Optus can’t advertise these services as an NBN substitute? Because it would be a lie. And it’s not like telcos don’t like to be free with the truth. You only have to see how often Optus was fined over their use of the word “Unlimited”. No one gets even close to those speeds in reality. But doesn’t “Upto 42Mbps” sound great but the reality is <10Mbps with regular dropouts.

    • Actually in an area with high amounts of users on the network due to limited ADSL say 25% you will be lucky to get dial up speeds if anything at all it is so bad I am seriously considering dial up to replace it :(

    • Agreed. You don’t get anywhere near those speeds, even when sitting right next to the tower. And the latency over 2/3/3.5/4G sucks big time – mostly due to the ordinary modem design targeted at end users.

  4. “currently offered speeds of up to 42Mbps down and 21Mbps up, with 84Mbps on the horizon”
    Why are there clauses to say Telstra and Optus can’t advertise these services as an NBN substitute? Because it would be a lie. And it’s not like telcos don’t like to be free with the truth. You only have to see how often Optus was fined over their use of the word “Unlimited”. No one gets even close to those speeds in reality. But doesn’t “Upto 42Mbps” sound great but the reality is <10Mbps with regular dropouts.

  5. i have both 3G wireless and ADSL. Previously I had 3G only at home due to a lack of ports. There is no way in hell would I ever consider a wireless only solution again. I welcome the NBN because at home I will have a no-compromise pipe to the net to do with what I wish.I then will step out the house and all the WIFI connections my 3G phones have will switch to the 3G network. Currently my family uses 8+ PCs in various guises around the house and we run out of bandwidth. I consider wireless as infill only. The path for this technology is best for people on the move and will provide the best speed possible the network can allow for a mostly seamless experience.
    latency in a fibre environment will be hard to beat as radio design takes into account re-sharing spectrum and does involve closing down unused virtual data pipes to efficiency. It will always take time to re-establish those links.
    Radio engineers working in the mobile field and technically aware colleagues fully support the NBN as they know the limitations of radio spectrum. The cost of NBN may be huge but future generations of Australians will thank us.

    • “The cost of NBN may be huge but future generations of Australians will thank us.”

      You sure about that – even if they are still paying for it?

  6. i have both 3G wireless and ADSL. Previously I had 3G only at home due to a lack of ports. There is no way in hell would I ever consider a wireless only solution again. I welcome the NBN because at home I will have a no-compromise pipe to the net to do with what I wish.I then will step out the house and all the WIFI connections my 3G phones have will switch to the 3G network. Currently my family uses 8+ PCs in various guises around the house and we run out of bandwidth. I consider wireless as infill only. The path for this technology is best for people on the move and will provide the best speed possible the network can allow for a mostly seamless experience.
    latency in a fibre environment will be hard to beat as radio design takes into account re-sharing spectrum and does involve closing down unused virtual data pipes to efficiency. It will always take time to re-establish those links.
    Radio engineers working in the mobile field and technically aware colleagues fully support the NBN as they know the limitations of radio spectrum. The cost of NBN may be huge but future generations of Australians will thank us.

  7. I am with marky_boi. I have a MAC protected wireless network at home with a desktop, 3 laptops, 2 iPhones and a tablet connected, going out through Optus Cable (which is infrastructure which seems to get ignored altogether in this debate – ADSL copper V wireless). Download speeds are OK for what I need, but could be better, and are getting cheaper. The iPhones and tablet connect to the home network when available (cheaper), then 3G when they are out (convenience). I have a Telstra 3G card for work, and the experience is altogther flakey – slow, patchy.

  8. @Frank Forsaker
    Please get over the idea that MAC filtering does anything to protect your WiFi against any halfway skilled beginner geek.

    I could sit outside your houe and sniff your traffic and then type a command into my laptop like:
    # ifconfig ath0 lladdr DE:AD:BE:EF:00:99
    having picked one of your MACs out of the air. Well all of the ones that were in use during the sniff.

    There’s a bit more to it but not much.

  9. Fixed line provides consistent speed (for ADSL, that means consistent for a given line…hopefully) and low latency. Wireless lives off to the side in the world of mobility and convenience. Fixed wireless sacrifices the mobility and convenience to provide semi-consistent speed and medium latency at a lower cost than fixed line.

    The “middle ground” between two fixed-line technologies must itself be a fixed line technology, like FTTN. The problem with FTTN is that it is half the cost of FTTH with a hundredth of the benefits. FTTN only makes sense if you are *certain* that there will never be any need for more than it can provide.

    >Why, if the two technologies will exist side by side, do all parties involved in the NBN process continue to make such a big deal out of it?

    The simple answer is that they do not. Coalition-biased journalists and politicians have made a big deal out of it. Actual experts and major providers (Telstra and Optus) do not. Telstra likes to brag about its mobile network, but it doesn’t pretend that it somehow replaces fixed line.

    >Why do there need to be clauses in the Telstra and Optus NBN contracts dealing with the issue?

    42Mbps NextG is comparable in speed to but less reliable than 8Mbps ADSL. The clauses are there to make sure that customers are not fooled into thinking that (mobile) wireless is anything other than a convenience.

    >Why can’t industry commentators stop pointing out the incredible growth of in wireless adoption, and the flatlining of fixed broadband?

    I don’t know. Maybe they confuse growth in popularity with growth in performance (in fact, an inverse relationship exists). Maybe they want to fool people into thinking that wireless is replacing fixed line. Maybe they have political motivations to do so.

    Both fixed-line and mobile data *usage* continues to grow.

    My family has four wireless connections and one fixed line. The fixed line connection’s usage is fifty times the sum of all the wireless connections (counting mobile phones, as the statistics always do). I suspect that this is fairly common.

    >Why does this theme come up again and again in the NBN debate?

    Same as above. Wireless *sounds like* an alternative, and it *sounds* cheaper. It would seem that many are simply obsessed with the idea of wireless, to the point where they ignore any of its limitations and grasp at every straw they can find (DIDO anyone?).

  10. Fixed line provides consistent speed (for ADSL, that means consistent for a given line…hopefully) and low latency. Wireless lives off to the side in the world of mobility and convenience. Fixed wireless sacrifices the mobility and convenience to provide semi-consistent speed and medium latency at a lower cost than fixed line.

    The “middle ground” between two fixed-line technologies must itself be a fixed line technology, like FTTN. The problem with FTTN is that it is half the cost of FTTH with a hundredth of the benefits. FTTN only makes sense if you are *certain* that there will never be any need for more than it can provide.

    >Why, if the two technologies will exist side by side, do all parties involved in the NBN process continue to make such a big deal out of it?

    The simple answer is that they do not. Coalition-biased journalists and politicians have made a big deal out of it. Actual experts and major providers (Telstra and Optus) do not. Telstra likes to brag about its mobile network, but it doesn’t pretend that it somehow replaces fixed line.

    >Why do there need to be clauses in the Telstra and Optus NBN contracts dealing with the issue?

    42Mbps NextG is comparable in speed to but less reliable than 8Mbps ADSL. The clauses are there to make sure that customers are not fooled into thinking that (mobile) wireless is anything other than a convenience.

    >Why can’t industry commentators stop pointing out the incredible growth of in wireless adoption, and the flatlining of fixed broadband?

    I don’t know. Maybe they confuse growth in popularity with growth in performance (in fact, an inverse relationship exists). Maybe they want to fool people into thinking that wireless is replacing fixed line. Maybe they have political motivations to do so.

    Both fixed-line and mobile data *usage* continues to grow.

    My family has four wireless connections and one fixed line. The fixed line connection’s usage is fifty times the sum of all the wireless connections (counting mobile phones, as the statistics always do). I suspect that this is fairly common.

    >Why does this theme come up again and again in the NBN debate?

    Same as above. Wireless *sounds like* an alternative, and it *sounds* cheaper. It would seem that many are simply obsessed with the idea of wireless, to the point where they ignore any of its limitations and grasp at every straw they can find (DIDO anyone?).

    • FTTN only makes sense if you are *certain* that there will never be any need for more than it can provide.

      Or if the cost of FTTH is such a leap that we cannot afford to make it. In general when (intelligent) commentators make the comment amount FTTN over FTTH this is what they are trying to say, that we can’t actually afford FTTH, especially on the scale of the NBN. I am a fan of the idea of reducing the FTTH footprint for this reason.

      My family has four wireless connections and one fixed line. The fixed line connection’s usage is fifty times the sum of all the wireless connections (counting mobile phones, as the statistics always do). I suspect that this is fairly common.

      This is quite common, except for non-technically motivated families. When the family is more technologically adept their data usage tends to go upward. A common argument is that most people are not technologically adept, hence their data usage will always be low. However, this argument tends to forget that the trend is for the lower-bound of data-usage to increase.

      Unfortunately for the moment, in terms of this argument at least, this lower-bound is able to be serviced quite well by wireless, which plays right into their argument, but I am of the strong opinion that this status-quo is not sustainable.

      Also, not important, however I do note that the counting of mobile phone data usage is not actually included in most stats I have refereed to, it is considered a separate and unique category to mobile broadband.

      Wireless *sounds like* an alternative, and it *sounds* cheaper. It would seem that many are simply obsessed with the idea of wireless, to the point where they ignore any of its limitations and grasp at every straw they can find

      It’s true, albeit it annoying, consequence of the current status-quo. For example, the NextG network only cost about $1b to build, and will likely only cost a few billion over it’s lifetime. The problem is that, like all wireless networks, the capacity of the network is rather limited, which can be easily seen by the fact quotas for wireless plans are usually in the order of 5 to 10 times more expensive as their fixed-line alternatives.

      It is quite clear that building a network to replace the wired networks is a huge expense, and intelligent commentators acknowledge this, but usually counter with the idea that higher and medium usage clients should be prepared to pay for high-capacity fixed line links in full, and not spread the cost across consumers that do not want it, i.e. only push fibre to those that want it.

      In principle this idea (only push fibre to those that want it) makes sense, in practice it is a fools errand. Aggregating capacity, even over the small scale of the last mile, is what makes home Broadband connections possible and affordable for the masses.

Comments are closed.