The Australian mobile spectrum

EM spectrum

While you’re wringing your hands about peak oil and fresh water shortages, add another woe to the list. The electromagnetic spectrum that gives us our communications and broadcast technologies is already crowded with services from radioastronomy to TV programming, and now we may have to pull more out of a hat.

After the blistering success of the iPhone has made Australians as hungry for mobile data as traditional market leaders like Japan and South Korea. In fact we’ve been downloading games, facebooking our friends and sharing our location so much it’s now affecting telecommunications policy. Towards 2020, An Australian Communications and Media Authority (ACMA) report, tells the story of our burgeoning love affair with data and how our infrastructure has to adapt to absorb its astounding growth.

The ACMA is the manager and resource allocator for the spectrum that delivers broadcast content and data over the airwaves (everything from mobile services, TV and radio to top secret military communications). And as the Authority knows firsthand, we’ll need more space on the spectrum to handle the unfolding mobile data explosion. Based on current estimates we’re looking at a projected 30-fold increase in mobile data between 2007 and 2015 and a whopping 500-fold increase by 2020.

Towards 2020 author Dr Andrew Kerans explains the physics of the data transmission limits we now face. “In 2007 about 380 MHz was available to communications carriers and we bought an additional 98 MHz into use more recently,” he says. “We’re planning to release about 100 MHz of the 700 MHz band in 2014 because of the introduction of digital TV. Together with another 140MHz of the 2.5 GHz band it’ll support the next generation — 4G or ‘long term evolution’ — of mobile phones. So there’ll be about 800 MHz available for mobile broadband in 2015, but spectrum planners think we’ll need about 150 MHz more. By 2020 they think we’ll need another 150 MHz on top of that, so that’s a total of 300 MHz we need to free up.”

Because like the minerals under our feet, the electromagnetic spectrum is a finite resource, and with bandwidth-heavy services like digital TV coming online to hog even more of it, the big question is where we’re going to find that 300MHz — even assuming the figure doesn’t change some time in the future.

Firstly, take any panicky ‘we’re running out’ stories on current affairs programs in the near future with a pinch of salt. Just like electronic fuel injection extended the usability of critically low fuel reserves in the 1970s, the word ‘finite’ is a movable feast.

“When you transmit a signal in a house it bounces off walls and comes to your phone by a number of different ways,” Kerans explains in one example of how changing technology is stretching the spectrum further. “In the old days the signals all interfered with themselves. These days the system is smart enough to carry different data on each path and combine it at the other end, so four different paths will give you four times the data. We can carry a lot more data within the same spectrum.”

Even better, we’ll soon have the infrastructure to let some broadcast services abandon the spectrum altogether. Since TV’s invention in the 1950s right up to the 1990s the airwaves were the only delivery method available, but Kerans thinks that could all change. “Broadcasting will probably end up on the NBN. They want 3D, multiple channels, interactive TV and all that. They’re not going to be able to do that on the radio systems they’ve got.”

It might sound disingenuous to talk about how the NBN is going to help free up spectrum in the age of tablets and data on mobiles, but it makes sense when Kerans talks about what he calls ‘mobile broadband delivered to much smaller cells fed by fibre’. “You’ll be walking around the house with your iPad and when you go out to your car it’ll switch from your wi-fi to Telstra or Vodafone, then when you get to work it’ll switch to your office system. It’ll flow seamlessly between them.”

But there are more players in the game than just the ACMA, and freeing spectrum real estate involves more than just government mandates. More efficient use of what we have will also come from market innovation. Kerans says he expects the telecommunications industry to deploy more infrastructure like improvements in transmitter technology — today we only need small antennas on the side of buildings or lamp-posts in disguise rather than the huge antennas that have dotted our landscapes since the early 90s.

The other driver is that your mobile provider couldn’t offer the grand final on your phone 15 years ago — the signal from the old generation networks was nowhere near strong enough to support that sort of bandwidth. But TV on your phone is just the pointy end of a flood of content. Telstra launched its 4G network in May, aiming to offer the service commercially all over the country by the end of 2011. New generation cells means an exponential increase in the amount of data in the signal at the same distance from the transmitter.

So far, so App Store. But in case you haven’t heard, there’s a digital storm of data coming, and it won’t have anything to do with Gmail, Facebook or apps for pasta recipes. The age of machine to machine communications is rising, and unless we want every washing machine, truck, hot water heater and power plant connected by wires, the spectrum is their only conduit to communicate. “Your fridge will talk to your smart meter and decide when to switch on and off,” Kerans says, “trucks will report their motor conditions and do vibration analysis on their bearings. That data’s going to continue to push the curve upwards.”

The spectrum the ACMA needs to find and free up exists in several different bands. If you’re really interested in the arcane physics you can read all about it in the report, but we still need the plan because despite all the gains and improvements mentioned above, we’re still going to come up short. “That’s spectrum we’re going to have to find,” Kerans confirms. “We’ve already taken improvements in technology and the increases in infrastructure into account.”

So the ACMA has its work cut out. As the Towards 2020 report says, getting spectrum that’s already home to other services will involve real commercial impacts. At every stage of the process it has to weigh up the agreements with incumbents who use it and the most benefit to the Australian public. Suddenly 2015 looks all too close…



The first commercial 1G mobile service was launched by NTT in Tokyo 1979, quickly followed by the Nordic Mobile Telephone (NMT) network throughout Scandinavia. Both systems led to the Advanced Mobile Phone System (AMPS), which was the basis for the early Australian networks.

1G was mostly analogue like the old-style PSTN phone network, so without any means to transmit data of any sort other than a voice signal that meant no apps, browsing, or even text messages (not the use of the word ‘mostly’ — even though the voice signal itself was transmitted in analogue like a radio station signal, the radio towers themselves were connected digitally).

Released in Australia in 1987, it’s the network we associate those old brick models with and operated at the 800MHz band.


It might surprise you to learn 2G is still used in many parts of the world. While it’s tempting to think ‘older equals obsolete’, the technology makes it more than adequate for some regions and uses.

The big difference from 1G was that the signal is digitally encoded all the way from one user to another, which means it can carry text and any other data formats as well as voice.

In a precursor to what we need to do with the spectrum today, 2G allowed more use through less bandwidth. Digital encoding meant the networks used less of the spectrum, so while packing more stuff into less space it also allowed for the rapid uptake of the mobile we saw in the early to mid 90s.

Digital’s other advantage was handsets that both needed and emitted less power, which meant smaller cells. That let us build the mobile transmission towers that sprung up across the world (the 1G equivalent would have been as big as skyscrapers) and improved device battery life.

But the lower power meant a weaker signal, so in order to take full advantage of the network you had to be comparatively close to a tower — the reason they’re so ubiquitous across our landscape today. And while a dodgy analogue signal merely means a few seconds of transistor radio-style static, a digital network will just cut you off.

A lesser-known upside of 2G was security. It was much easier to assign a unique phone number to a device, where analogue systems made it possible to ‘clone’ devices, a practice which led to cases of fraud. You can also apply security algorithms to digital signals, which makes 2G harder to eavesdrop or pick up with radio scanners.

Later iterations of the 2G networks you’re probably familiar with from the little readouts on old phones are GPRS (General Packet Radio Service), CDMA (Code Division Multiple Access) or EDGE (Enhanced Data rates for GSM Evolution).


3G gave the humble mobile its wings. Launched in 2001 (in Japan again), it was a United Nations-led standards specification called the International Mobile Telecommunications-2000. It (eventually) allowed for expanded voice signal technology, mobile Internet access and video and TV signals for calls and downloads over a then-prehistoric download rate of 200kbps.

Australia was host to the first demonstration of 3G in the southern hemisphere after tech-savvy nations like Japan and South Korea launched commercial services. m.Net switched it on in 2002 in the 2100MHz band, in a totally different range from the 2G networks. The first commercial service, Three (hence the name) was launched by Hutchison Telecoms in March 2003.

But 3G is about location as much as just bandwidth Even though the new generation of devices would be doing a lot more for us based on where we are, who knew that in 2001? As a result, the noughties saw a flurry of applications for standards in telemedicine, location data and video streaming for teleconferencing or watching movies.

As the mobile and wireless chips and under-the-hood processing of devices has grown, they’ve become able to download even faster. The promises of 21 or 42Mbps by our carriers is marketing-speak (you’re more likely to get about 1-3Mbps in an average Australian capital city), but the capability’s there.

Like all data pipes, it depends how many other people are accessing the network — you might have got a better speed on your 2G phone in 1997 simply because you were the only one in your street using the network. Add to that the vagaries of natural and man-made geography and interference from other services and mobile networks are still too patchy for life and death enterprises. Even in the tech-smart US the fastest network (AT&T) managed a maximum of 9.41 Mbps but an average of only 2.24Mbps during recent tests.

Two of the names you know from the 3G era are GSM (Global System for Mobile Communications, originally Groupe Sp├ęcial Mobile) and HSDPA (High-Speed Downlink Packet Access)


Aside from the usual constraints on mobile networks mentioned above in geography and user demand, going from 3G to 4G is going to be a bit like going from a Goggomobile to a Lambourghini. The new speed requirements in the ITU Radiocommunication Sector (ITU-R) specifications are for 100 Mbit/s for high mobility (if you’re in a train or car) and a whopping 1 Gbit/s if you’re standing or walking.

Why so high? Read any story about smartphones and you’ll see a line about how it’s a full-featured computer in your pocket. Smartphone sales are set to overtake those of PCs worldwide next year, and we could conceivably be witnessing the first steps in the end of the PC era because of the processing power carry around in our handheld devices.

Throw in the popularity of tablets thanks to the success of the iPad and we could be using handheld devices for everything in the foreseeable future — a mammoth amount of data filling the sky above our heads when we’re all unplugged from the wall.

WiMAX (Worldwide Interoperability for Microwave Access) and LTE (Long Term Evolution) have been out since 2006 and 2009 respectively and though they’re sometimes co-opted by the marketers, they’re not strictly 4G yet — their peak bitrates of 144 and 100 Mbps don’t cut it. Watch out for WiMAX2 and LTE Advanced instead.

Because of the always-on nature of mobile data today, one of the hallmarks of 4G is seamless roaming, connectivity and handoff between networks to make your mobile experience constant no matter what networks you’re switching between.

4G will also take advantage of better data transmission protocols that allow for MIMO (multiple input, multiple output). It allows the same band of the spectrum to be used for data flowing both ways because transceivers at either end can disentangle the information, giving the spectrum another efficiency boost.

That in turn will pave the way for the smooth deployment of IPv6. In the machine-to-machine communication age there’ll be a lot more than just us and our iPads, and we’re already running out of unique identifiers because of the 32-bit IPv4 system.