George Malim finds that the devices needed for LTE tests and trials are entering the market, while smartphones will arrive in six to eight months 
 
 
 

With LTE networks rolling out in many markets, devices are currently confined to single mode, single band dongles. However, LTE will operate across different frequencies from market-to-market and devices will need to interoperate with 2G and 3G technologies as well as offer the features of smartphones that users demand.
Traditionally, availability of endpoints lags behind development of network technology and that causes delay from trials, testing and deployment perspectives. The problem stems from the nature of device design. The chipsets that need to be developed to support new network technologies need to wait until network standards are finalised before the design work can begin and it’s a time consuming process once started.
“Handsets are always the critical part in these sorts of developments because you’ve got to factor in a chip design cycle so devices always end up coming later than the infrastructure side,” says Gordon Aspin, chief executive of Cognovo, a spin out from ARM Holdings that uses software-defined modem techniques to overcome the problems of early availability.
The concept behind the company’s software engine is to allow chipset hardware to be constructed early and then enhanced using software to suit deployments. Even with this approach, the chipset design workload is still substantial so, while it provides a means to accelerate development, it doesn’t eradicate the lag between network and device development.
For Stephane Daeuble, head of global LTE product marketing at Motorola, device availability is always a limitation in new network technology deployment, but he is encouraged by the steps being made with LTE endpoints.
 
 
WiMax chip vendors leading
 
“The delay is limiting but it is no different to previous technologies,” he says. “The difference this time is that non-traditional suppliers are at the forefront. That’s because LTE and WiMax are similar from a chipset point of view and it is those vendors that are leading development.”
“From the networks’ point of view devices never come early enough,” adds Daeuble. “Standards need to be very, very firm before chipset design can progress but adaptation can now be based on software so you can modify as you go. Development is also accelerated with LTE because, whereas devices have typically been maybe six months behind the network, the WiMax chipset makers in particular are actually further along.”
The first step is to address the needs of operators that have live deployments of LTE. “Our perspective on LTE is that we are now in the phase of execution of the plan in line with what has been announced,” says Enrico Salvatori, senior vice president and general manager of Qualcomm Europe. “The most concrete deployment is Verizon’s and we are part of that from the device side and are working with the major infrastructure vendors that are working with Verizon. Testing is ongoing, trials are ongoing and we can definitely comment positively. From our perspective, we’re on track with the delivery of our communications devices.”
 
 
Frequency challenges
 
One of the issues LTE faces is that operators will deploy using different frequencies. For example, deployments in the US will work in the 700 megahertz band while the majority of European deployments will be in the 2.6 gigahertz band.
“LTE frankly is very complex not just in itself, because the standard isn’t too difficult, but in the multiple bands and legacy nodes that you have to deal with in the same devices,” says Aspin. “There are 11 to 13 frequency bands that devices will have to work across and that depends on where you sit in the world. There are different views as to what the ideal band is.”
Daeuble thinks much of the work required to address it has already been done. “Frequency is not so much of an issue,” he says. “If you take a milestone of developments to date, there are three or four GSM bands, three or four UMTS bands with wifi and Bluetooth also required so there are ten or 12 bands supported already. These frequencies are already built in to phones so I don’t think it’s much of an issue.”
Even as development of multimode devices and smartphones continues at full speed, LTE devices are available in limited numbers. “Today you have trials in TeliaSonera and others and Samsung is providing some dongles into that market,” says Aspin. “Those are single mode terminals to validate LTE, it’s a logical approach for today.”
Daeuble agrees. “We need devices all the way through the process of testing networks internally,” he says. “Operators today are really looking for testing networking and want tens, hundreds and in some cases thousands of devices. I’ve just returned from China and I was able to play with early commercial LTE dongles and we were getting pretty decent experiences, we’re pretty close.”
However, he emphasises that large scale deployments of LTE are only happening in very select markets and therefore large volumes of fully featured devices aren’t needed just yet. “Sweden is probably the most advanced market,” he says. “They’ve gone commercial and I’ve praised them for doing that but it is a limited deployment around Stockholm and a few other cities.”
 
 
Commercial launch
 
Salvatori also sees Sweden as leading the European charge. “There’s a lot of activity in terms of timescale,” he says. “Operators in Sweden, in particular, are very aggressive and are targeting very late 2010 or early 2011 for full commercial offerings.”
Regardless of how far we are from the mass market, single mode dongles won’t be enough to deliver on the promise of LTE network capacity.
“The industry has learnt a lot of lessons from the debacle of 3G and its launch,” says Aspin. “The NGMN — Next Generation Mobile Network — grouping of operators is trying to help terminal producers understand what the needs will be so at least there’s a common requirement rather than individual operator demands. At least that’s the idea.”
The challenge is for device and chipset makers to get to a price, form factor and power consumption that will be attractive to users.
“To get to a product people will use, you need multimode,” says Aspin. “Early devices tend to be bigger, higher in power demand and costlier.”
Salvatori agrees. “The challenges have been to develop a solution that has lower power consumption and size,” he says. “The footprint has to be the same size and the device has to have multimode capability because, in our vision, 4G will a complementary technology to existing 3G, especially in very dense urban areas.
“In the 3G market at the moment there are really no 3G-only devices. We think that in the short to medium term the technologies are complimentary so multimode capability is almost a mandatory feature.”
 
 
Single mode to multimode
 
Daeuble also sees that development path. “As devices come through, they start with single mode, single frequency capability and six months later become multimode,” he says. “The stack is already built in the chipset, so there’s effectively everything operators need in the chip. It just isn’t used at the moment.”
Dongles are the obvious first wave of LTE endpoints. The form factor presents no challenge and it is anticipated that the first users will be connecting via netbooks and laptops. “The technology is available now,” says Salvatori. “The first devices will be USB modems and that first wave will focus on netbook or laptop connections and we see that as the first demand coming from operators.”
With development aided by the ability to use software to fine tune chips, the first commercial deployments active and a wider understanding of users’ demands than previously seen in new network roll-out, it looks increasingly unlikely that device availability will present a barrier to the mass market roll-out of LTE.
The work that remains to be done is feasible to complete in the timescale allowed before vast numbers of users demand fully-featured devices.
Aspin is confident that in spite of the apparent lack of current LTE devices they won’t hold back usage of LTE networks. “Device availability probably won’t limit LTE,” he says. “Technologists always overestimate the speed of technology adoption and devices will be available at the end of this year and there will be sufficient there to satisfy initial demand.
“It will take two to three years to build into a mass market and I don’t see availability as the critical factor. I think LTE is very much a positive move for the whole industry. 3G was too much of a technology push but it’s very clear people want broadband on the move and LTE will meet the requirements for people to do that.” GTB