Level Up: LTE Release 12

Every cellular technology is a work in progress—even today, more than 20 years after the first call was made, there are ongoing efforts to squeeze further performance out of GSM networks. So while the progress of new standards is expressed in terms of achievements like network launches, device availability and subscriber uptake, there is no fixed destination—no point at which the work can be deemed complete. These milestones represent only the culmination of individual projects and the industry always has one foot planted firmly in the future.

With LTE renowned as the fastest-growing technology in the industry’s history it follows that its evolution will be similarly accelerated. Indeed the first LTE-Advanced deployments—including early moves into carrier aggregation—are already underway in leading markets as ground-breaking operators look to push the technology into new performance territory.

But the process by which these advances are realised is not one of straightforward, technical problem solving. How to improve the standard, which technologies to promote and which paths to follow, is a matter of impassioned and exhaustive debate. This is largely because most of the companies carrying out the technical standards work are vendors looking to monetise their R&D by positioning their solutions for inclusion in the standard. A year ago 3GPP started work on Release 12, which is expected to be frozen at the end of this year. According to Dino Flore, chairman of 3GPP RAN, LTE has historically moved forward in tworelease cycles, starting with the base features in R8 and R9 and through the IMT-Advanced 4G fulfilment achieved with R10 and R11. It is too early to say for certain, he says, which enhancements currently under discussion will make it into R12 and which will be held back for R13. A presentation to 3GPP’s RAN June 2012 workshop from Korean LTE leader SK Telecom provided a good overview of the issues that need to be addressed. SKT said that it expected its spectrum resource to improve by a factor of three to reach 920MHz by 2020. In line with this it wants to improve spectral efficiency by a factor of six and resource reuse by a factor of 56.

Flore says that four top-level requirements emerged from that workshop, which featured presentations from more than 50 companies, including 14 operators. Work is being directed towards:

• Increases in capacity to enable operators to cope with the forecast explosion in network traffic;

• Improvements in energy- and cost-efficient operation;

• Support for diverse applications and traffic types; and

• Improvements in data throughput and user experience.

A huge variety of technologies are being explored to address these four broad needs. There is, of course, more than one way to cut the cake. Ericsson groups its approach to LTE evolution by technology area, according to Joakim Sorelius, LTE product manager at the Swedish vendor. Work across Releases 11, 12 and 13 is segmented into four areas, he says. “The first is related to spectrum, the second is the hetnet and the third area is antenna solutions. The fourth, which is perhaps not so important for the network, is the device side. One aim here is to specify a simpler type of chipset that does not have such high processing capabilities for M2M type devices. Then there is work on even more advanced receivers for even faster devices, and also on device to device communication,” he says. Ericsson’s groups are easily matched to the takeaways from last year’s 3GPP workshop. Either way, says Dino Flore, capacity is the biggest issue.

“The most frequent requirement that was made during last year’s workshop was a need to address this network traffic explosion. Many companies quoted the expectation that we will have 1,000x growth in traffic over the next ten years. This was the main concern for both operators and vendors.”

So how is this demand for capacity best met? For Thorsten Robrecht, head of portfolio management at Nokia Siemens Networks, carrier aggregation is the key. “I’m seeing more and more that this is the big thing for operators,” says Robrecht, “and it’s where we’re putting the most effort.”

Carrier aggregation was specified in R10 and the early deployments we will see later this year are based on that first specfication. But technology is already moving on, Robrecht says, offering increasing benefits—particularly to smaller operators that have ended up with fragmented spectrum portfolios due to financial restraints.

R12 Features (Click To Expand)

“With the coming technology these varied spectrum landscapes don’t matter,” he says. “We are not yet at the point where you can aggregate everything from 800MHz to 2.6GHz but it is coming. And at least the combination of the small chunks into one bigger pipe is really starting to happen with Release 12.” Joakim Sorelius reports a similar enthusiasm for carrier aggregation at Ericsson. With the kind of demands likely to be placed on networks in the medium term standardisation work must focus on solutions that derive truly substantial capacity gains, he says. And the biggest gains to any system are achieved by adding more spectrum. Ericsson is pushing carrier aggregation hard with Australian operator Telstra, which expects to launch a 900/1800 aggregation solution later this year. The problem with carrier aggregation is that spectrum is scarce, or not always readily available. Telstra has just secured valuable holdings at 700MHz and 2.5GHz, for example, but these licences don’t become available for commercial use for almost another 18 months. So what are the alternatives for operators looking to exploit existing spectrum holdings? “With existing spectrum the best answer is the use of small cells,” says 3GPP’s Dino Flore.

“We introduced hetnet support in Release 10 and 11 but with Release 12 we’re going one step further, towards hyper-densification. To improve spectral efficiency you need to augment the number of antennas in the network. You can do that with MIMO or by increasing the number of small cells, and the latter emerged as the most reasonable approach.” Ericsson’s Sorelius singles out a particular Release 12 feature, the “soft cell”, by way of example. It is a concept that sees the addition of a new carrier that “allows you to add data channels without adding the overhead in terms of control signalling. That’s a Release 12 feature we believe in very much.”

But Mike Wright, executive director for networks and access technologies at Telstra, who is working closely with Ericsson on carrier aggregation, points out that small cell deployments are also tied very closely to spectrum holdings. Despite being one of the world’s most advanced LTE operators, Wright says Telstra doesn’t have the scale or resource to dedicate to active participation in 3GPP. Instead it contributes through GSMA and vendor partners and he doesn’t share Flore’s unqualified enthusiasm for hyper-densification.

“I think it is important, and this year we will do our own hetnet trials, looking at stadiums and central business districts. But in a place like Australia the distance between people makes it somewhat impractical to hyperdensify in some areas.” For Wright there is as much if not more value in the near term in investigating emerging antenna technologies like beamforming and higher order MIMO. “Ultimately the choice is between densification and getting the most out of the towers that you already own. To be honest I’d rather get the most out of the towers we own first, before we go building tens or hundreds of thousands of small cells. Because they will have a life cycle and an operational cost attached.”

Sorelius isn’t convinced. Ericsson isn’t actively pushing higher order MIMO schemes as commercial solutions because “device vendors have put their multi-antenna soliutions quite far into the future—if ever,” he says. Current solutions do not justify the investment, as they offer more limited returns; 10 – 20 per cent in the best cases by his assessment.

The firm has demonstrated 8×8 MIMO in a microcell product but the UE was “like a suitcase,” he says: “It’s not really feasible from a practical perspective.” For testing and learning and enabling advances on the component side there is value in exploring these technologies but commercially viable solutions are too far off to be considered now, he says, adding: “The motivation of [vendors] that do not have a very large footprint is to try and push these more esoteric advanced features as a differentiator.”

This is not Sorelius’ only potentially provocative observation. One of the most hotly contested debates within 3GPP, he says, surrounds the potential impact of multi-vendor hetnet installations. “There are some players who would very much like it if there was no [negative impact] created by putting a small cell from one company into a macro domain supplied by another,” he says. “But there’s quite a lot of evidence showing that the type of tight co-ordination that you need is done on such a small timescale that you really need to be within the same eNodeB to make it happen— preferably within the same base band HW board or between tightly coupled boards.” Related to the push on small cells is the need to further enhance interworking between LTE and WLAN. Previous work has been done at the system level but R12 will see radio interworking taken “to the next level,” says Dino Flore. Key requirements include the ability for operators to steer traffic on and off WLAN on a truly dynamic basis, with the process improved in terms of efficiency and speed.

Clearly there is no shortage of ideas for the evolution of the LTE standard into the next decade. Nor is there any lack of focus; demand for wireless broadband services is heaping pressure on operators and their suppliers to deliver whatever improvements they can muster. Perhaps time is the only thing in short supply—and with Nokia Siemens Networks and T-Mobile, to name two, now talking publicly about 5G—there is no easing off the throttle.

LTE D2D – Who needs the network?

Device to device (D2D) communication has emerged as one of the key working areas for 3GPP Release 12, moving beyond STG RAN into the 3GPP System group. Suggested applications include off-net peer to peer communication, backup for emergency services in disaster situations and proximity/discovery applications.

Proponents of the technology—Qualcomm is one of the most vocal, having proposed its LTE Direct solution for inclusion in Release 12—argue that D2D technologies address key obstacles to the mass adoption of proximity and discovery services. Privacy concerns, battery drain and unwarranted network load have all acted as a drag on enthusiasm to these services, for both the operator and the consumer.

With LTE Direct, applications on user devices do not log the user’s location and status in a central server, they simply broadcast either a need or some kind of service offer over a directly allocated resource in the LTE uplink in a series of pulses that Qualcomm has termed “expressions”.

“These are 54 or 128 bit identifiers and each device might emit mutliple expressions,” says Prakash Sangem, director of technical marketing at Qualcomm. “The expression is so quick that there’s no need for messages, it’s all broadcast. The apps will match what someone’s looking for with what’s being sold or offered.”

It may have the ring of a Location Based Services conference presentation from the late 90s but Sangem argues that these types of service can only be made to work if scaled up. There are going to be very few matches in an area the size of Times Square where there might be several thousand devices, he says. “You have to scan a huge amount of services and devices to find out what is relevant to you.”

The suggested commercial benefits to operators are familiar: advertising revenue, push marketing and so on. But using licenced spectrum for the discovery process should help keep operators in control of a game that is currently dominated by OTT players, Qualcomm suggests. But encouraging device to device communications, essentially off-net, might create issues in terms of revenue collection for core communication services. Meanwhile network interoperability is another, familiar obstacle to the kind of scale that Sangem argues LTEDirect will enable.