Telcos are saying 4G will help Australia keep up with growing demands for mobile bandwidth. But like cloud, 4G has become a nebulous term, leaving users unsure as to what it actually is. We delve into the mysteries of 4G and investigate how it will help mobile phones keep up with customers' hunger for data.
The International Telecommunications Union (ITU) is the body responsible for deciding what speeds and standards make up a new mobile generation. 1G referred to analog mobile communications during the 1970s, while 2G refers to digital systems developed at the end of the 1980s, such as CDMA or GSM, which are still used by telcos from around 170 countries across the globe today — Optus, Telstra and Vodafone here in Australia.
But after 2G, it started to get complicated. While development on the 3G standard we use today began in 1992, it wasn't until 1999 that the ITU officially declared some developed technologies as meeting the "IMT-2000" 3G standard of offering faster voice and internet services with seamless global roaming. The first 3G networks began being deployed in Asia in 2000 and 2003; Australia got its first taste for 3G with the launch of Hutchison Telecommunications' 3.
The 3G standard was a serious step up from 2G that it became a victim of its own success. Telstra last reported that 70 per cent of its customers are now using 3G.
When the first iPhone arrived in 2008, it brought along with it an enormous increase in demand for data. Australian mobile networks found themselves unprepared for the onslaught as customers used their new phones to bombard networks. Service became patchy and customers complained, with Optus being one of the most criticised at the time, although Vodafone felt the heat last year when its coverage didn't meet demand. While Telstra's Next G network has so far maintained the reputation of being the superior network in Australia, many have said it has begun to struggle under the weight of its 12 million customers, particularly in the CBD areas of Sydney, Melbourne, Brisbane and Perth.
In order to cope, Optus has now invested millions upon millions of dollars in upgrading its networks, particularly in high density population areas. Vodafone has invested $1 billion this year alone in upgrading its network after its infamous outages at the end of 2010.
Despite investments, the telcos' troubles aren't over. Globally, mobile broadband traffic is expected to increase from 5 exabytes this year to 35 exabytes by 2014. Yet, short of putting a cell tower on every street, 3G upgrades can only go so far, so finding a new technology that can deliver more data to increasing numbers of customers in crowded metropolises is the only solution.
Therein lies 4G.
What is 4G?
4G, or what the ITU refers to as IMT-Advanced, is classed as being a significant jump from existing 3G technologies in a number of ways, while still being able to be used around existing networks.
The term '4G' itself remains undefined.
International Telecommunications Union
The ITU expects to provide detailed specifications of IMT-Advanced (4G) technologies in a new recommendation in early 2012, but for now, the basic requirements a service needs to meet to be considered as 4G are:
- A high degree of functionality, flexibility and efficiency
- Compatibility with existing networks
- Capability to work with other radio access systems
- Availability of user equipment worldwide
- Worldwide roaming ability
- Peak data rates of up to 1 gigabit per second.
The ITU told ZDNet Australia that no technology currently deployed around the world met that definition.
"The term '4G' itself remains undefined, but it is being applied by operators to the forerunners of IMT-Advanced technologies, such as [Long Term Evolution (LTE)] and WiMax, and other evolved 3G technologies," the ITU said.
"However, IMT-Advanced, incorporating the most advanced technologies currently defined for global wireless mobile broadband communications, is considered as '4G'."
So while the ITU has not yet officially declared any one new technology to be 4G standard, it is aware of the global marketing trend to label new network technologies as 4G. This has occurred not only in the US, where we have seen telcos such as Sprint label WiMax networks as 4G, but also in Australia, where both NBN Co and Telstra have added the 4G label to their LTE networks.
But why do this when the ITU hasn't declared LTE to be 4G?
Because long-term evolution is on the path to 4G.
As we saw with the long gap between ITU investigations of 3G in 1992 and the actual deployment of 3G networks in 2000, it can take a long time for new technologies to be deployed commercially. 4G is no different.
Industry experts in Australia tip that the next level of LTE, commonly known as LTE-Advanced, will likely be the first incarnation of 4G.
The LTE standard was first conceived in 2004 by the global organisations that formed the 3rd Generation Partnership Project (3GPP), which originally came up with GSM. The 3GPP had wanted something to cope with the massive use of video on mobile networks. It came up with a standard that uses an Evolved UMTS Terrestrial Radio Access Network (E-UTRAN) and a new packet core called an evolved packet core.
In naming the standard, 3GPP set a number of goals for the technology to achieve, including:
- A peak download rate of 100Mbps and a peak upload rate of 50Mbps within 20MHz;
- Very low latency
- Compatibility with existing 3GPP networks
- High spectral efficiency.
These specifications alone don't meet the ITU's definition of 4G, according to Huawei Australia's director of LTE solutions Nick Bromhead. In 2009, however, 3GPP proposed to the ITU that the 10th LTE release and onwards would aim to meet the 4G standard and would be called LTE-Advanced.
While reaching the maximum speeds of between 100Mbps and 1Gbps were important, Bromhead said that spectral efficiency would be key.
"If you look at the spectral efficiency of a HSPA+ system ... it's around 1 megabit per second per Hertz, and LTE is around 1.7 and LTE-A is about 2.2 megabits per second per Hertz."
Ericsson Australia's general manager of strategic marketing, Kursten Leins, said that while there was very little difference in spectral efficiency between traditional HSPA technology and LTE, one of the main advantages to LTE was that it supports much wider bandwidths than traditional HSPA technology, which makes it easier to produce devices to work with the technology.
"The complexity in the devices really reduces because you're not combining four different carriers, four different spectrum allocations in the single device," he said. "You're actually using a much wider bandwidth in the first place."
Leins said this will make devices cheaper to produce, and will ultimately drive up adoption of the technology.
Unfortunately, this is where LTE branches again, meaning that making head or tail of who's offering what even more complex. LTE is essentially broken up into two types — Time Division Duplex-LTE (TD-LTE) and Frequency Division Duplexing LTE (FDD-LTE). Leins said that the main difference between the two methods were how they utilised spectrum.
"So for FDD, you have basically separate blocks of spectrum, one dedicated for uplink, one dedicated for downlink, and they're actually separated in the spectrum band," he said. "With TDD you have a single block of spectrum ... It's timing, effectively.
"It's just a different way of handling the data. All cellular networks today tend to use frequency separation because it tends to give you better coverage," he said. "It comes down to an efficient way of building networks."
All the carriers are looking to LTE to ease network congestion, but Leins said it won't replace 3G networks, which he said will be around until the end of this decade, at least.
"The current networks, 3G and 2G networks, they have sufficient capacity to handle voice traffic. So the reason most markets are deploying LTE is the data volumes," he said. "The underlying voice capacity on 3G and 2G. Voice grows, but it's fairly incremental."
"It's just simply more subscribers signing up to data services, and there's more of them. So that's what's driving it," he said, adding that voice services on LTE were a while away, but any LTE phones that will be released will likely fall back to 3G in order to make voice calls.
For Telstra, part of the reason for launching LTE was about driving down the cost of providing data.
It doesn't make sense to choose a different technology path when you don't have backwards compatibility.
Ericsson Australia's Kursten Leins
"The cost per bit is significantly lower on an LTE network as opposed to a HSPA+ network. So it's all about the efficiency of carrying data on the network. LTE is just a far more efficient technology in terms of the cost base. And remember, for us to be able to maintain margins in that market, we really need to be driving down costs on the network," Telstra CEO David Thodey said recently.
Meanwhile, WiMax, the alternate next-generation mobile technology to LTE, seems less likely to gain a foothold in Australia. The former coalition government had plans for a $1 billion OPEL project that was going to deliver broadband to regional and rural Australia through WiMax technologies. However, that was canned by the current Labor government to make way for its National Broadband Network. Vividwireless had also been heading down the WiMax path, but now looks to be switching lanes into LTE.
According to Leins, the reason that telcos have picked LTE is because it's on the ongoing evolution path of the same standards group that does 2G and 3G networks, which would require much less work (and cost) for telcos to upgrade their networks.
"It's very hard to change, it doesn't make sense to choose a different technology path when you don't have backwards compatibility," he said.
"You've got certain core networks, systems, sites, cell spacing set up for a particular mode of operation. As an operator if you choose to ignore that and start fresh, it's a big financial decision," he said.
Instead, carriers are choosing to keep their older networks active and interoperable with new networks. Vodafone has done just that in Australia, by replacing all its existing Ericsson 2G and 3G network cabinets with one Huawei SingleRAN cabinet that will deliver both 2G and 3G services and support LTE coverage.
What's stopping us from having LTE today?
In Australia, it seems, deployment of LTE networks is being held back by one factor: there's no spectrum.
One of the reasons we're seeing 1800[MHz] from Telstra and VHA: it's really the only option they have.
Ericsson Australia's Kursten Leins
Carriers are, not so patiently, waiting for the Australian Communications and Media Authority's auction of "waterfront" 700MHz spectrum, which is being freed up by the move from analog to digital television. This is scheduled to occur in 2012. In the meantime, Telstra has forged ahead with its LTE plans, deciding to use the 1800MHz spectrum that it currently has (used for 2G service, which are required less and less) for services in Sydney, Melbourne and Brisbane, until more spectrum becomes available. Vodafone has also committed to using its 1800MHz spectrum for the new technology.
"One of the reasons we're seeing 1800[MHz] from Telstra and VHA: it's really the only option they have," Leins said.
"1800 is quite a good band because it's used globally and was used globally for GSM, and that's why Telstra's been quite successful in getting others on board with 1800."
In addition to being the areas where people are more likely to adopt LTE right away, the CBDs are more likely to have more towers, as would be required of the higher spectrum frequency. To ensure the future of LTE on any spectrum, the devices will likely be multiband, meaning that once more spectrum is made available, it will be able to switch over as the LTE network is expanded.
And it will be the 700MHz spectrum from the auction where the main LTE game will be played. This spectrum is viewed as ideal for providing data services, and is already in use for the largest deployment of LTE in the world — Verizon in the United States. Huawei's Bromhead said that the lower frequency spectrum is ideal because it can cover a wider distance and has greater building penetration.
"All the operators are waiting with great anticipation to get their hands on that spectrum. If you double the frequency then you need almost three times the number of base stations to cover that area," he said.
"The economics of the operators are very much in favour of having the lower frequencies because they don't have to build as many sites."
Another reason behind the delay in moving to Long Term Evolution, according to Leins, was the availability of fibre backhaul.
"One of the problems is, you can't just put in a whopping great cell site and not get the traffic out of there, otherwise you'll have a bottleneck," he said.
This is one area where Telstra had an advantage on its competition, he added.
"Telstra's in the fortunate position where they have something like 90 per cent plus of their sites fibre connected," he said. "So they've got very high capacity backhaul."
It may be one reason as to why Vodafone looked to secure fibre backhaul to its mobile base stations as part of its recent agreement to get on board the NBN as a retail service provider.
Optus was a late comer to LTE announcing last week it would launch FDD-LTE services in 2012. CEO Paul O'Sullivan has stated that he believes devices will be at a mature level by that point and there will be enough consumer demand for LTE. Bromhead said it was early days for LTE devices, but the ones that Australia will get will likely support a variety of spectrum bands as well as traditional 3G networks.
"What you'll actually find is that the original LTE devices, they will support LTE on 1800, 2600 and then they will support 3G at 900, 800 and 2100," Bromhead said.
Once the auction has gone ahead all carriers will likely be charging head long into the new mobile standard and onwards to Long Term Evolution. So it won't feel that long until Australians get to taste the faster speeds that 4G will have on offer.