LTE: a new horizon?

15 February 2011 |

With the momentum behind LTE, has WiMAX been left behind? Caroline Chappell investigates whether there will be a clear winner in the race to 4G.


When Verizon Wireless launched its LTE network in 38 US markets at the beginning of December 2010, analysts believe it sealed the fate of its rival ‘4G’ technology, WiMAX. "WiMAX may have been the better standard, but the industry decides which technology will win," comments Bengt Nordström, CEO of research consultancy, Northstream. "WiMAX supporters underestimated the size of the established (GSMA-based) ecosystem," Nordström suggests. They also failed to understand mobile operators’ appetite for a single mobile network technology standard to help them lower the cost of upgrading their networks and to stimulate demand. To cope with the tsunami of mobile data demand washing over them, wireless operators need to give their subscribers order of magnitude increases in connectivity speeds. 4G technology provides the means to do this: LTE and WiMAX both promise subscribers a truly mobile version of the fixed broadband network, with speeds that can easily outstrip those available from fixed network providers.

However, if mobile operators are to make money from offering their smartphone and tablet-owning customers vast increases in bandwidth, they need to build out 4G networks with the lowest levels of capex they can get away with. Three years ago it appeared that WiMAX held the cost advantage card, with suppliers predicting that LTE equipment would be up to 10 times more expensive to deploy. But when operators of the stature of China Mobile and Verizon Wireless declared for LTE – "we decided to jump on the (GSMA) 3GPP bandwagon as we saw that the long-term future of mobile networks would be in that camp," says Lindsay Notwell, Verizon’s executive director, LTE development – most network equipment vendors saw the writing on the wall and swung behind the standard too. And the outlook only got sunnier for such suppliers when, not to be left behind in the 4G race, AT&T announced it would roll-out an LTE network this year – albeit at a more leisurely pace than Verizon Wireless. Today, Nordström believes: "WiMAX prices and business cases are often worse than for cellular 4G technologies."

Global support for LTE

"The fact that LTE has global support behind it was a key factor in our decision to adopt it. CDMA is North American-centric and doesn’t enjoy the advantages of global economies of scale," explains Ed Chao, CTO of MetroPCS. MetroPCS pipped Verizon to the post as the first US operator to launch an LTE network in September 2010. "Global competition drives down the cost of handsets and network equipment and increases innovation. We saw that LTE was going to be the foundation for a lot of great advances and better prices."

Not surprisingly, Christian Hedelin, head of product and portfolio marketing, radio and OSS, for Ericsson’s business networks unit, agrees with Chao’s and Notwell’s analysis. "The success of GSM is largely down to the economies of scale it has brought into play, especially on the terminal side, which has made GSM networks available to more than four billion people worldwide," Hedelin says. "Low device prices widened accessibility and this is why operators are looking at a 3GPP standard." Ericsson was one of the first network equipment vendors to discontinue its investment in WiMAX, a decision the company "doesn’t regret" as LTE looks set to become the dominant 4G technology.

But WiMAX is not going to disappear just yet, in the opinion of Ofer Karp, president of Alvarion’s wireless business division. WiMAX is the more tried and tested technology which has already been deployed in nearly 600 commercial networks in 149 countries, according to the WiMAX Forum. Implementors range from established mobile operators, such as Sprint and Korea Telecom to a host of alternative players entering the mobile market. Karp says that WiMAX has been as popular with "ISPs, using WiMAX as a replacement for, or extension to, their DSL footprint in underserved regions, such as rural areas, as with operators such as Clearwire and KDDI, which have built networks from scratch in dense areas of population." Even Julian Grivolas, principal analyst, networks and technologies at Ovum and a long-time WiMAX sceptic, admits: "Clearwire has done great things, starting from nowhere to cover 100 million people with an entirely new technology and interesting devices." And Clearwire is enabling Sprint’s 4G successes too: Sprint 4G is present in 71 US markets and introduced its 16th and 17th 4G devices at CES 2011 in January. 

WiMAX demand still increasing

WiMAX has a firm foothold in the market and its early adopters are synonymous with 4G in many people’s minds. Karp says that WiMAX demand is still increasing and is fairly equally spread worldwide, in both developing markets, especially Africa and Asia-Pacific, and developed markets, such as the US, where mobile broadband penetration is growing at a rapid rate and new spectrum and network capabilities are needed to support it. "We expect the WiMAX market to grow by 10% to 15% over the next few years and stabilise at a peak of $2.5 billion," Karp says.

"WiMAX’s key advantage today is time to market," points out Ozgur Aytar, research director, Pyramid Research’s broadband and media research practice. "In markets where players already own spectrum, the sooner they get to market, the faster they get their money back and WiMAX deployments make sense, even if they don’t enjoy the same economies of scale as cellular technologies." In contrast, "despite strong momentum for LTE, it will take a long time to commercialise the technology," Nordström asserts. The road to LTE adoption will be a bumpy one, not least because the spectrum situation is even more challenging than it was for 3G. "The launch of 2G nearly 20 years ago was the best spectrum situation the world ever had. The timing was well co-ordinated and a number of markets launched at the same time," Nordström notes. "We lack the same level of harmonisation with 4G." Countries are moving at different speeds to auction spectrum that could be used for 4G services and there is more fragmentation: "The US is using different spectrum bands to Europe and Japan is different again," Nordström points out. "Technically a single device can support multiple spectrum bands but this makes them more complicated and expensive to build." 

Longevity of HSPA

LTE will also be affected by the longevity of HSPA, which currently enjoys enormous operator support round the world. Indeed, Nordström recommends that "the bulk of a mobile operator’s investment should go towards HSPA and HSPA+ at present because it’s a mature technology and where all the devices are. Smartphones function better with HSPA while the important role for LTE over the next couple of years will be in the dongle/replacement ADSL market." Grivolas agrees: "HSPA has a long life ahead of it and will be the dominant technology until 2015. But we don’t see HSPA and LTE as competitors: they are complementary. If you are an operator with HSPA+ and enough spectrum, then you will be able to sustain traffic growth for the next two to three years and can go slow on LTE. Then you can come later to the market when LTE has matured, it has a larger ecosystem and lower cost." Those operators making a big splash with LTE today are those like Verizon Wireless, which have reached a dead end with their current (CDMA) technologies, analysts suggest, or TeliaSonera, one of the few PTTs not to win a 3G licence in its home market, which has been forced to co-operate with its largest competitor over the past decade.

Although WiMAX will continue as a niche market for some years, "what the market needs is an honourable transition from WiMAX to LTE," Nordström suggests. After all, both technologies share the same foundation – OFDM (Orthogonal Frequency Division Multiplexing), a technique for transmitting digital data over a radio wave – so bringing the two together in a multimode base station is not seen as an enormous task. However, since WiMAX uses TDD (Time Division Duplexing) spectrum and initial LTE products use FDD (Frequency Division Duplexing) spectrum, this means developing a TDD version of LTE. Network equipment vendors are minded to do this in any case, to meet the requirements of TDD spectrum holders such as China Mobile, which want LTE.

Such transition technologies are already under development, if not already here: in October 2010, Huawei launched the world’s first WiMAX/LTE SingleRAN solution and is poised to trial its LTE TDD implementation in Clearwire’s network. Meanwhile, Sprint has announced ‘Network Vision’, a development in conjunction with Alcatel-Lucent, Ericsson and Samsung, which will result in multimode base stations able to support Sprint’s entire range of spectrum bands, including spectrum used by CDMA and WiMAX. Sprint expects to spend $4 to $5 billion re-equipping its network but to make operational savings of $10 to $11 billion over a seven-year period as a result.

Mobile operators are making their technology bets and the largest and most advanced are well on their way to becoming all-IP players. Migration to 4G technologies will help them cater for the relentless growth in demand for mobile broadband while developments such as Sprint’s Network Vision and technology standardisation will help them to contain costs. Nevertheless, the 4G technology race is just a starting point: mobile operators still face a large challenge in settling the business case for 4G.

What counts as 4G? 

As Verizon’s Notwell points out: "There is a tonne of noise in the marketplace about 4G because definitions are so fuzzy." Wikipedia’s self-evident definition is that 4G is the fourth generation of cellular wireless standards, following on from 2G and 3G. But according to Wikipedia, certain speed criteria need to be met for a network to be called 4G. Peak download speeds for a 4G network must be 100Mbps for high mobility communication, such as from trains and cars, while pedestrians and stationary users should be able to receive 1Gbps. None of the so-called 4G networks currently deployed qualify, according to this definition, whether LTE or WiMAX.

Indeed 4G speeds – an average Verizon user in a loaded network will get 5-12Mbps while Clearwire promises 3-6Mbps and bursts of over 10Mbps – look slow compared to 3.5G HSPA speeds. "HSPA is now a mature technology and all vendors can deliver 21Mbps," Nordström points out. "A few have launched 28Mbps, several are trialling 45Mbps and 84Mbps HSPA is on the roadmap. So to add to the confusion, some operators, such as T-Mobile, argue that they have just as much right to call their HSPA networks ‘4G’ as current LTE and WiMAX operators do."

In December 2010, the ITU decided to ‘allow’ current generations of LTE, WiMAX and HSPA networks to use the term ‘4G’, even though they don’t deliver on the specified download speeds at present. ITU recognises that these networks nevertheless offer a "substantial level of improvement" over today’s 3G technologies. The ITU’s pronouncement also appears to acknowledge the fact that mobile operators are constrained by their spectrum ownership and the state of play of technology development as to the speeds they can offer. According to Ovum’s principal analyst of networks and technologies, Julian Grivolas, this is a complex problem which makes comparisons between technologies and operators difficult. "Verizon is using 10MHz FDD, so the download peak rates should be around 50Mbps max," Grivolas says. "But with 20MHz FDD, download speeds can be around 100Mbps in a commercial network and the standard allows up to 150Mbps, or even 173Mbps in some cases. Add in 4x4 MIMO (smart antenna technology) and the standard will allow up to 300Mbps, although due to the cost and complexity of this technology, this won’t be implemented for some years." Grivolas adds: "If you compare theoretical speeds, Verizon can offer 50Mbps LTE compared to T-Mobile’s 42Mbps HSPA."

Latency is not mentioned in Wikipedia’s 4G ‘definition’, but here there is clear blue water between ‘4G’ LTE and ‘3.5G’ HSPA. LTE’s round-trip airlink latency is around half of HSPA’s. Hedelin believes that "delay is underestimated as a factor in poor user experience and reducing latency is becoming more important than the download speed," especially in some of the more advanced applications operators expect to deliver over a 4G network.

For some time, it has been widely agreed that a 4G network is one that natively supports IP, able to deliver any IP application, including VoIP (a standard for Voice over LTE, or VoLTE, is presently being defined). In this case, contenders for the 4G title are limited to WiMAX and LTE, with their flatter, all-IP network topologies helping to reduce latency and equipment costs.


One to watch: the US market 

To a mobile subscriber, the technology that underpins their ‘4G’ experience is irrelevant. Subscribers simply want access to their services as fast as possible, wherever they are and at the lowest cost. Regardless of whether a mobile operator implements WiMAX, LTE or HSPA+, it will need to differentiate itself on the factors that matter to a customer: coverage, price, experience, devices and applications. The rest of the world may find it instructive to watch the US market in this respect. The US is one of the world’s most advanced 4G markets, with every 4G technology represented there and a number of 4G operators starting to go head-to-head. How such players differentiate themselves will set a benchmark for other markets.

Coverage will be critical: Sprint believes that it has significant 4G coverage advantage for the time being with 71 markets compared to Verizon Wireless’ 38 markets, but Verizon has aggressive plans to roll out its LTE capability. And Verizon 4G devices will be backwards-compatible with its 3G network so that applications such as video calls will maintain connectivity if they roam out of LTE coverage, but reduce resolution to adapt to 3G speeds.

Unlimited connectivity will be the norm: Sprint and MetroPCS are setting the bar high with their ‘all-you-can-eat’ price plans for 4G mobile data services. Sprint believes its 4G data plans are currently better value than any other carrier but MetroPCS expects to give Sprint and Verizon a run for their money as the best value supplier of mobile data plans in its 14 metro markets.

Customers will look for the best speeds, lowest latency and least loading. These factors all affect the customer experience and 4G subscribers will soon learn which operator provides the best experience for their money. LTE operators are likely to have the lowest latency – Verizon Wireless is promising round-trip airlink latency of 30ms – though there is not much to choose between 4G providers in terms of speeds. Subscribers today may be delighted with the lack of congestion on such new networks but this situation is unlikely to last. On the other hand, as MetroPCS’ Chao points out, new networks have been engineered to deliver speeds well in excess of market needs today.

It’s all about the device. 4G users will expect to be able to use a single device regardless of the extent of their provider’s 4G coverage and to use a growing range of devices – laptops, iPads, smartphones – with a 4G network. Verizon Wireless has ensured that 4G devices will be backwards compatible with its 3G network but so far, its LTE network will only support USB modem users: those wanting to use smartphones will have to wait for a limited number of devices until the middle of the year. Sprint’s headstart in the 4G market means that it has the largest ecosystem of devices and operators implementing HSPA can also provide device choice to their customers. In contrast, MetroPCS currently only supports one Qualcomm device which ‘talks’ voice on MetroPCS’s CDMA network and supports browsing on its LTE network, although it expects to release a portfolio of Android smartphones in 2011. There are snags with LTE devices at present, analysts point out, for example, on MetroPCS’ current device, users can’t receive calls and browse at the same time.

Operators need to partner for premium applications. Both Sprint and Verizon Wireless have set up third-party application developer programmes to encourage the creation of new and attractive applications for their 4G networks. Notwell reports that there has been "a tremendous response" from its 6,000-strong Verizon Developer Community to its LTE roll out. Smaller operators, such as MetroPCS, may struggle on the applications front. MetroPCS has developed its own social networking and premium ‘video snacking’ applications for its Qualcomm device, but these may be overtaken by over-the-top applications as Android smartphones become available.