Should the NBN worry about wireless?

Should the NBN worry about wireless?

Summary: Outspoken chief executive of Australian internet service provider Exetel, John Linton, has been clear that he considers that wireless kills the NBN Co's fibre-to-the-home proposition. But could mobile broadband really rain on the NBN's parade and turn it into a failed field of dreams?


Outspoken chief executive of Australian internet service provider Exetel, John Linton, has been clear that he considers that wireless kills the NBN Co's fibre-to-the-home proposition. But could mobile broadband really rain on the National Broadband Network's (NBN) parade and turn it into a failed field of dreams?

"The average person needs a 100Mbps internet connection about as much as they need to have their arms amputated," Linton said at the beginning of this year.

It was a typical attack by Linton on a project that the Federal Government sees as today's equivalent of the Snowy Mountains Hydro Scheme, but that Linton sees as "a field of dreams" dreamed up by "a total wanker" — Prime Minister Kevin Rudd.

It's this kind of rhetoric that has led many in Australia's telecommunications industry to discount the importance of Linton's views.

However at the core of Linton's attack on the NBN in January was a kernel of insight that is currently worrying many in that industry. It's not the kind of worry that will lead to the NBN project being cancelled. It's going full steam ahead and the NBN Company has started looking for contractors to start digging up Australia's streets and lay fibre-optic cable around the nation.

But it's the sort of worry that is currently coming up in conversation at industry events again, and again, and again. In short, could the potential of mobile broadband technologies become such a powerful reality over the next few years that they make it impossible for the NBN to ever achieve a commercial return?

Channeling Linton

Linton has emerged as one of the main critics of the NBN, and although we couldn't get him to do an interview for this article, his daily blog has become a hotbed of debate about the project, and one that is sometimes surprisingly based on hard numbers. For example, just this Monday he pointed out that the Australian Bureau of Statistics is due to publish today its half-yearly report on usage of broadband services by Australian consumers.

Linton has been looking forward to this report for some time. As he wrote in January 2010, comments from the major mobile carriers — VHA, Telstra and Optus — have indicated the take-up of mobile broadband in the last six months of 2009 was much faster even than the 162 percent reported by the Australian Communications and Media Authority for the 12 months previously.

Linton said at the time that it was likely that the number of mobile broadband connections in Australia must be approaching three million — which if true, he said, was "an extraordinary technology take-up rate — by far the fastest Australia has ever seen".

It's true.

The ABS figures released today show that as at the end of December 2009, there were 2.838 million mobile broadband connections in Australia. That figure is a staggering 814,000 connections higher than it was in June 2009 — a growth of 40.2 per cent in six months alone. And the June 2009 figure was itself up 48 per cent on December 2008.

At the same time, the number of fixed-line broadband subscribers in Australia barely changed in the last six months of 2009.

There were 4.193 million ADSL subscribers at the end of 2009, compared to 4.171 million in June 2009 and 4.176 million in December 2008. And there were just 935,000 cable and fibre subscribers at the end of 2009, according to the ABS, compared with 931,000 in June 2009.

Broadband Statistics

Broadband Subscriber Statistics (Credit: ABS)

The statistics are clear. Australians are adopting mobile broadband solutions in record numbers. It's no wonder that at times the networks of the mobile carriers have struggled to keep up with the demand. The usage is absolutely exploding.

This leads to Linton's next question.

"If you were going to 'bet the company', on rolling out a new residential communications infrastructure over the coming five to 10 years, what would you use as the technology base?" he asked in a blog post yesterday, pointing out many European telcos were currently focused on mobile options instead of fibre — such as the incoming rush of technologies based on the Long Term Evolution (LTE) standard, which Telstra and Optus are also trialling locally.

Well, in some ways the stats say it all. Wireless is grabbing market share as fast as it can. And it's easy to see why.

"It's safe to say that Australia is in a very good position globally when it comes to mobile broadband," says Peter Rossi, the chief technology officer of Huawei Australia, the local division of the Chinese networking vendor.

"Not only do all Australian operators have national HSPA networks which are consistently being upgraded and expanded, but vividwireless has also launched Australia's first WiMAX network — a clear indication of the growing consumer demand for high-speed wireless broadband services."

What are those subscribers doing?

Some 673,000 of those mobile broadband customers belong to VHA, the mobile telco formed out of the merger of the local operations of Vodafone and Hutchison (which operates the 3 brand). The company's head of Broadband Products, Klaas Raaijmakers, says the company classifies mobile broadband users — those who user a 3G USB card tethered to their laptop or desktop — into two categories.

The first class, "substituters", refers to those who have dumped their fixed-line broadband connections and now just use mobile, while the second class, "complementers", use both, complementing their fixed-line access with mobile broadband when they're on the road.

Three years ago, when mobile broadband was taking off, he says, the first group was the dominant purchaser of mobile broadband solutions.

This was partially caused by the lack of real competition in the fixed broadband market at time time, he says, but also people liked the versatility and flexibility of mobile broadband and its "instant gratification" nature — you just walk into the retail shopfront of a mobile carrier, buy a USB modem, and walk out with it working instantly, instead of having to wait 15 days for ADSL to get connected.

Customers had started to get excited about mobile broadband offerings when the price point had hit $29 to $39 a month, he said.

However, there has recently been a switch in the type of customer that VHA has been signing up. Now more users are "complementers". These users, Raaijmakers says, see the internet like the oxygen that we breathe — they need to be connected 24 hours a day, seven days a week, no matter where they are.

Complementers, according to the executive, tend to use mobile broadband for those purposes it is most often associated with — web browsing (especially search and internet banking), email and so on, with social networking becoming an important category in the past couple of years.

Substituters, in comparison, tend to use those services as well, but also more high-end services online, such as uploading photos and watching streaming video from sites like YouTube.

But Raaijmakers points out that there is still another class of broadband users — those who do use the internet as a tool like the others — but also increasingly as an "entertainment substitute". Instead of watching TV, he points out, many users download shows or watch streaming video online. They don't have a cable TV service like Foxtel. They have the internet for that.

The executive admits VHA doesn't currently cater for this class of users.

"We're quite clear on that," he says. "If you look at our plan structures, they end at seven or eight gigabytes [of download allowance]." Using the internet as your entertainment source would likely require more than that — many ADSL plans go into the tens of gigabytes or even beyond 100GB.

It's the need to serve this growing class of users that forms one of the key arguments for why wireless will not significantly impact the NBN.

One of the reasons for this is the problem of shared bandwidth — unlike ADSL infrastructure, where each user gets their own pathway back to the backbone internet networks through telephone exchanges, bandwidth on mobile phone towers is shared between users who generally connect sporadically, rather than constantly downloading streaming media over a period of hours.

David Kennedy, a research director with analyst house Ovum, says the company does believe that wireless broadband will be in competition with fixed in certain segments, particularly for young people and for those that only require low-bandwidth applications.

However, he points to the long-term demand for so-called 'triple-play' bundles — voice, video and data in one telecommunications package — as something that wireless can't do that well.

"Wireless is not a great medium for emulating that service package because of the video component," he says in a recent interview. As mobile young people get older and settle down, he says, they often start to sign up for existing household packages like Foxtel — "young people don't stay young forever".

"Triple-play is typically associated with wireline broadband — it's fixed connections which have traditionally had the raw throughput necessary to deliver video, voice and data," says Huawei's Rossi.

"That said, there's been consistent growth in mobile video services alongside the more obvious boom in mobile broadband — many operators see burgeoning demand for mobile video as a key reason for network upgrades. vividwireless is a perfect example — the vivid network is built for mobile data, but also seamlessly runs the engin VoIP service and TiVo content is being offered quota-free."

Kennedy says about 8 per cent of households in Australia currently use only wireless broadband. At most, he says, this might double over the next few years and be enough to make a difference to the business case for NBN fibre to the home services — if not as much of a difference as some, such as Exetel's John Linton, might believe.

However, Kennedy says, when you look deeper into that business case, it's actually some of the other questions around what people want and are willing to pay for that are interesting.

For example, he points out that many Australians only have 1.5Mbps broadband speeds currently as they're on ADSL1 technologies or cheaper, throttled back ADSL2+ plans. And they're happy with those speeds, as they're only using their connection for online services such as internet banking.

"When you hear people complaining about broadband, they're pretty much a minority," he says. "Most people are happy with 1.5Mbps. What this shows is that at the very least, the uptake of high-speed services is going to be a very slow process."

The release of iiNet's tentative pricing plans for the early stage NBN rollout in Tasmania this month fuelled the debate around what people are willing to pay for what speeds — customers will be required to pay $129.95 for the full 100Mbps NBN speeds, while slower plans will be more on par with ADSL2+ pricing.

This ties into Raaijmakers' statement that customers really got interested in mobile broadband at the $29 to $39 monthly subscription mark.

"Working out what retail prices will be acceptable is very difficult," says Kennedy, pointing out that global benchmarks haven't been very helpful, with StarHub's fibre service in Singapore only attractive a few thousand customers despite offering services for more than a year.

Queensland-based software developer Sam Moffatt has been using Optus mobile broadband as his main broadband connection for the past few months because he can't get copper cable laid to his house, as it's on the same block as another house which is already connected.

He pays $100 a month for an Optus prepaid connection, which he says has its ups and downs, "being wireless". "Sometimes speeds will be ordinary and sometimes good," he says. He gets about 6GB of download allowance included and has received speeds of about 2Mbps at times, although it can slow down sometimes.

Asked whether he would be attracted to getting a NBN connection when the fibre is rolled out in his neighbourhood, Moffatt says "yes and no". On the one hand, he likes the mobility the Optus mobile connection gives him — he's recently been travelling a bit and has been able to take his internet connection with him.

On other hand, he can see the upside of faster speeds, higher download quotas and competitive pricing from the NBN, although he notes "the NBN would have to add significantly more and be significantly cheaper". And the ability to download high-quality view — such as through the ABC's streaming iView service — would be attractive.

The official view

The broad view espoused by proponents of the NBN project and many others in the industry is that fixed and wireless broadband services are actually complementary in many ways, rather than being competitive.

"The growth in wireless services does not have to be at the expense of fixed broadband," Communications Minister Stephen Conroy said earlier this month. "At a more technical level, wireless and fixed broadband technologies are complementary."

Conroy and others have pointed out that mobile broadband depends on having good fixed-line backhaul connections to phone towers, and that mobile would benefit in this way from the roll-out of the NBN, through more affordable backbone traffic. However, Telstra said several weeks ago that it had already invested a large amount in bringing fibre backhaul to its towers.

NBN Co chief executive Mike Quigley and head of industry engagement Christy Boyce have both pointed out in public speeches that bandwidth demands overall have continued to grow rapidly over the past several decades. NBN Co anticipates that by the time the NBN build is finished in almost a decade's time that Australia will need all the bandwidth it can deliver.

Even some wireless chiefs — such as the Seven Network's Ryan Stokes, a key figure in the development of the company's vividwireless subsidiary, which is rolling out wireless solutions around the nation — agrees Australia will need both fixed and wireless solutions.

"Just as wireless could not be a substitute for the NBN, the NBN is not a competitor for future demand for wireless-delivered high-speed data," he said in a recent speech. Huawei's Ross, taking the global view, agrees.

"There is consistent investment in both wireless and fixed-line services globally, and I think most operators would agree that the services are complementary," he says.

"Just as operators are looking at LTE and WiMAX to deliver advanced mobile broadband networks worldwide, that doesn't mean they've shunned fixed lines — the level of global activity in deploying fibre-to-the-home or fibre-to-the-node shows demand for fixed-line services is alive and well."

Will wireless services cannibalise the NBN? Only time will tell.

Topics: Broadband, Networking, NBN

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
  • NBN would have to be truly universal to be a contender in the home space. Without compulsory connections to existing dwellings, that will take years.

    For small business, it is another matter entirely, because high upload speeds is where ADSL and most wireless services fall down. Telstra is the only one offering relatively low-cost multi-mbps uploads, and that is in wireless (usually better than most other telcos' download speeds). Sending emails with attachments without such speeds really hobbles communication.

    ADSL2(+) speeds drop off a short distance from the exchange, so for most it is an empty promise. NBN should have the distance issues (at least to anywhere the same extent).

    The NBN may well enable a wholesale movement of people being able to work from home because their VPN connections can finally support the bandwidth to work as if they were on the company LAN, but without all the freezeups they have now!

    As surveys have shown, wireless is popular among:
    - renters = immediate connection and no transfer fees
    - shared housholds as each can have whatever they want without affecting others.

    We went all wireless when we moved city as we could have low-cost internet the next day in the motel to be able to search for jobs and houses, and email.
  • This is a simple problem of shared medium vs non-shared medium.

    Radio networks are inherently broadcast at the physical level. Two radio devices active on a network must share out the available bandwidth between them. Adding a second client device to a network with one base station and one client device will slow down the first client device. Overheads increase with the number of devices too, so at 1000 devices each device won't receive 1/1000 of the speed, it'll probably get less.

    Activity rates moderate that, of course. It's unlikely that all 1000 devices will be actively using the network at maximum capacity all at once. Even so, adding more users will significantly affect existing users.

    Anybody who's been to a concert and tried to use their mobile phone or mobile Internet has some idea what this is like.

    This means that even if 4G services could achieve their theoretical 300Mbit/s or so, if there's any significant number of subscribers and especially if they're using it more than sporadically, the network can quickly become oversubscribed and consequently overloaded. Current 3G and 4G networks are very lucky to hit 1/10th of that overall throughput across all users, with individual users lucky to top 5Mbit/s, so they're just not viable for wide-spread deployment for heavy Internet use.

    By contrast, a FTTN or FTTH network doesn't generally share the access medium. Even in a PON setup where it's possible that some spurs may share bandwidth, the overall capacity is so high that the network can guarantee a high minimum of uncontended bandwidth per endpoint. It doesn't suffer as you add subscribers.

    Of course, both wireless and FTTN/FTTH need a fast enough uplink and backbone, or everything's bottlenecked there. But that's not specific to the delivery technology; it applies no matter what you use.

    The point is that wireless may be suitable for "patching the holes" where FTTN/FTTH can't reach or hasn't yet reached, and may also be good for users who need roaming Internet and/or are quite light users. It's simply not realistic for everyone in the street to use wireless Internet because it's a shared-access medium that just doesn't have the capacity.
  • I would point John Linton and anyone who takes what he says seriously to the article named "Jammed"in New Scientist 30/10/2010.
    Basically, rather than wireless being a competitor to a fiber NBN, a fiber based NBN will be the saviour of wireless.
    The reason is simple, wireless technology has improved to the point where it is pushing the theoretical limits.
    Current 3G has managed to squeeze 1 data bit into 1Hz of bandwidth. That is, one 1 or zero transition into one positive to negative waveform transition. LTE squeezes it slightly more because the positive to negative transition is actually a positve - zero - negative transition so you can, in absolutely perfect conditions, get 1.5 bits into each transition.
    Sadly that's it folks. Short of magic or Dr Who's Tardis, the inside (data signal) cannot be bigger than the outside (the carrier bandwidth).

    Currently wireless traffic is doubling each and every year (actually, a little faster than that). On current estimates, places like New York will go into total meltdown sometime in 2013. Other metro areas will follow shortly after.
    Building bases isn't the solution. If there are 10,000 bases this year, in a perfect world, we would need to build 10,000 more next year, to provide the same data rates per user, 20,000 new bases the year after, 40,000 the year after that.

    Last year in this country, most 3G mobile bases got a second frequency. This year many were expanded to 6 sectors (6 transmitters) effectively 2 bases per site. Next year some extra 21MHz carriers may be brought on line. But thats it, there is nothing else in the pipeline. After that data will need to be rationed.

    There IS a solution however, and the New Scientist article discusses it. Femtocells.
    Basically they are a small mobile base in a box that can be installed into each and every house.
    Each and every house that is that has a reliable, high speed data connection.