The great broadband race – A US perspective
By Phil Marshall
After several false starts, consumers are now embracing mobile broadband and as a result, service providers in the US are scrambling to economically address market demand. While each service provider has its own approach to delivering mobile broadband, most are touting their solution as 4G. Verizon is aggressively pursuing Long Term Evolution (LTE), AT&T is essentially ’slipstreaming” Verizon’s efforts and plans to enhance its High Speed Packet Access network with (HSPA+). T-Mobile lacks spectrum to deploy LTE and is pursuing a strategy to aggressively pursue HSPA+. We believe that it is also investigating partnership prospects with Clearwire. Clearwire is also supporting Sprint’s mobile broadband initiative, initially with WiMAX which we believe will migrate to FDD-LTE in select markets, and ultimately TDD-LTE once it becomes available.
While there are marginal performance benefits with LTE over that of HSPA+, the major benefit for players like Verizon, AT&T and Clearwire is to leverage new radio spectrum resources. We estimate that Verizon and AT&T will increase the aggregate capacity of each cell site by 40 and 60Mbps for each LTE 10MHz carrier deployed. This contrasts with the 50 to 75 percent performance gain that we estimate between optimized LTE and HSPA+ networks.
As HSPA+ and LTE take hold, the need for multimode and multiband devices and infrastructure will balloon. A player like AT&T will require GSM/EDGE, UMTS/HSPA/HSPA+ and LTE modes traversing 700, 850, PCS and AWS spectrum bands. Verizon will require similar support across the same spectrum bands with CDMA 1X, 1X Advanced, 1xEVDO and LTE support. From an infrastructure standpoint, equipment vendors like Alcatel Lucent, Ericsson, Motorola, Nokia Siemens Networks, Huawei and ZTE have made significant strides with software defined radio (SDR) technologies to support multimodality. Solutions that deliver integrated multiband capabilities remain a challenge.
Since LTE technology has essentially been optimized for data services and leverages an end-to-end IP architecture we do not expect its wide use for voice telephony services any time soon. CDMA based technologies like UMTS and 1X are better suited to voice relative to OFDMA based technologies like LTE. In addition, it will take many years before end-to-end VoIP is capable of achieving comparable performance to traditional circuit switched voice over the air interface. We believe that both UMTS and CDMA 1X will prevail in the absence of viable substitutes for many years.
The race for mobile broadband supremacy amongst service providers will accelerate in 2011. As this occurs, we can expect significant experimentation in terms of service offers. Players will struggle with ballooning traffic and in the face of challenging economics will accelerate efforts to introduce network intelligence and traffic offload solutions. We believe that AT&T is ahead amongst its peers in developing these solutions.
Femtocells – A business case for free?
By Andrew Mitchell
Monday evening’s ShowStoppers at CTIA Wireless 2010 lived up to its billing as being an event showcasing “hot companies with cool products”. With around 30 vendors present there were lots of cool products, ranging from applications, to accessories as well as a number of innovative new security solutions. But what really caught my eye on the way into the busy room was actually not a product but rather the presence of an association - Femto Forum.
While Femto Forum itself isn’t a “hot company” and doesn’t produce “cool products” it did have a very cool message for the many press and media folks who attended the event. Femto Forum’s mission in wireless focuses on driving standardization, regulation and interoperability of femtocell technology. The organization also focuses on providing support for the marketing and promotion of femtocells as solutions for the wireless industry.
I had the pleasure of speaking with Femto Forum’s enthusiastic chair, Professor Simon Saunders. When asked “what’s new?”, Saunders provided some highlights of two interesting and compelling studies that advance the inclusion of femtocell solutions in 4G networks. Outlining the first study Saunders explained that the forum’s research found that the addition of femtocells to the network allows femtocell users to consistently receive much closer to the headline LTE/WiMAX data rates than those connected to macrocells.”
On the surface, this might be reason enough for a consumer to consider the purchase and use of a femtocell. To an operator, including support of femtocells could contribute to improving user experience while at the same time reducing the demand for increased macrocell infrastructure. The study suggests that an operator with 10 million LTE or WiMAX subscribers with a deployment of femtocells to 10% of its subscribers could expect to see a return of over 10 times for its investment.
But that’s not all. Saunders went on to explain that another study commissioned by the forum supported a model where an operator could actually provide early LTE/WiMAX subscribers with a free femtocell to deploy in a home or office so as to create the best possible experience indoors. Depending on variables, the model suggests that funding for the free femtocell would come from the opportunity to delay anywhere from 4 to 10% of the operator’s required infrastructure build.
The business cases would seem to be compelling enough. Looking down at the table I see sample devices, including one branded for AT&T. There are also devices branded for France’s SFR and Japan’s Softbank. In the case of AT&T, it’s a 3G device and perhaps one that will provide that operator with some relief from the on-going challenges it’s had in supporting the seemingly unquenchable demand for bandwidth presented by the increasingly more prevalent iPhone. Regardless of whether its an operator’s current challenge or future opportunity, it appears obvious that femtocells have solid, and perhaps even a “free” 4G future.
AT&T goes with Ericsson and Alcatel-Lucent for LTE
By Caroline Gabriel
However greedily Huawei has been biting at the mobile vendors’ heels in their European heartlands, the US door remains firmly slammed in its face. AT&T has announced its LTE vendors, a year almost to the day since Verizon Wireless did the same thing, and has chosen exactly the same RAN suppliers - Ericsson and Alcatel-Lucent.
This will be a huge relief to both suppliers, which have been feeling the pressure from Huawei in the first round of LTE trials and contracts. The Chinese giant has moved beyond its old reputation for competing purely on price and shown itself able to challenge even in cutting edge technologies. And with the firm overtaking ALU and Nokia Siemens in the overall mobile infrastructure rankings last quarter, the traditional players need to ensure their bugbear does not convert too many of its early LTE trials into major deployments.
Breaking into the US is tough for Huawei, largely for political reasons, and so far it only has 4G deals with Clearwire for WiMAX and cableco Cox. This leaves north America, which promises to surpass even Japan in early adoption of 4G technologies, a surprisingly clear run for Ericsson. Once the Swedish giant’s weak spot, the region now accounts for 12% of group revenues, following the acquisition of CDMA, LTE and GSM assets, and the huge outsourcing deal with Sprint. Providing LTE kit for both the big two cellcos gives it a good springboard to dominate this market going forward, pushing more entrenched local players like Motorola aside - even if LTE, as Ericsson has said itself, will not generate meaningful revenues for a couple more years.
Alcatel-Lucent, the incumbent in north America thanks to Lucent’s huge CDMA base, will be even more relieved to have secured part of the AT&T contract. Despite its win at Verizon, it has so far been overshadowed in European and Asian LTE by Ericsson and Huawei, and badly needs to ensure that the decline in CDMA revenues and margins will be offset by new sales, most urgently in the US.
AT&T will start field trials later this year and plans initial commercial deployments in 2011, a timeline that has been brought forward slightly, probably to address fears that the cellco would fall too far behind Verizon in mobile broadband capability. These fears have been particularly acute given the failure of its 3G network to support the mobile data explosion, much of it driven by its iPhone exclusive, in key urban areas.
Although the carrier is upgrading its whole network to HSPA, it is only moving to the 7.2 Mbps iteration, while many operators in other parts of the world are implementing 14.4 Mbps or even HSPA+ (21 Mbps or 42 Mbps) for areas of high capacity requirement. AT&T will certainly have some catching up to do, if it is to retain Apple’s favors, support new data-driven devices like the iPad and netbooks, and compete with Clearwire and Verizon LTE in key metro markets.
Initial roll-outs will focus on these city-based regions of high data demand. Despite their early timescales compared to European carriers - most of which will not start deploying until 2012 at the earliest - Verizon and AT&T are expected to confine their roll-outs to fairly limited areas of high population density for at least a couple of years. Verizon will also use LTE for some rural access requirements, in its 700MHz spectrum, whose long range is strong for covering sparse populations and getting signals indoors (though problematic for dense urban hotzones, where the revenues will lie). AT&T is also expected to use 700MHz spectrum acquired in the 2008 auctions, and also to overlay some of its 3G networks with LTE, linked to a common core.
The firm played down its timelag behind its arch-rival, hinting that Verizon would roll out before there were devices available anyway. By contrast, John Stankey, CEO of AT&T Operations, said the AT&T schedule “aligns with industry expectations for development of LTE technology and widespread availability of equipment and compatible LTE mobile devices”.
The carrier also said that “as part of the supplier agreements, 3G equipment delivered to AT&T by the suppliers starting this year will be easily convertible to LTE, enabling A&T to upgrade existing equipment and software rather than install entirely new equipment in many cases”. This will be one factor in the choice of Ericsson and ALU, both of which currently supply AT&T’s HSPA network, though the cellco said it had tested equipment from other providers too. But they will support a parallel strategy for 3G and LTE, which Stankey was keen to contrast with Verizon’s more abrupt migration path between CDMA and the new platform.
“AT&T has a key advantage in that LTE is an evolution of the existing GSM family of technologies that powers our network … As some competitors move away from their existing investment in niche 3G platforms, we are able to efficiently and quickly move toward LTE while enhancing our existing 3G performance and providing access to a strong ecosystem of customer devices,” he said in a clear swipe at Verizon - a swipe from which ALU will be keen to distance itself, since one of its key selling points to its CDMA base is the ease of migration it offers with its multimode strategy.
Financial terms of the deals with Ericsson and ALU were not disclosed. The clear loser in the US market is now Nokia Siemens. With the top three cellcos secured in terms of 4G RAN, NSN only has a contract with Verizon Wireless for its LTE IMS (IP Multimedia Subsystem), shared with ALU, having dropped out of the Clearwire supplier list at an early stage. It will be hard for the firm to maintain the brave face it put on Verizon’s choices a year ago, when then-CEO Simon Beresford-Wylie commented: “This is their first wave of build-out - we’re right at the beginning of this. I look forward to being there in the second wave.”
Auctions must be decoupled to release brakes on mobile broadband
By Caroline Gabriel
Auctioning licenses in several bands at the same time appears to be a good idea, and one increasingly favored by regulators round the world. Rather than selling each set of frequencies in sequence, and often with no over-arching strategy, governments can offer bidders the flexibility to acquire the right mix of spectrum for their business models.
So India decided to offer 3G and WiMAX spectrum at the same time, allowing operators to create plans that combined 3G for coverage with 4G for data intensive zones and fixed broadband. The authorities argued this would be a better model than one based on a single technology and frequency, both for commercial return and the expansion of services around the country. In the UK, a very different market, the idea was to auction the two bands earmarked for 3G expansion and 4G - 2.6 GHz and digital dividend around 800 MHz - simultaneously. Again, this would allow operators to make a clearsighted judgement on what balance of frequencies they needed to balance wide coverage with urban capacity, and to value the bands accordingly, rather than in isolation.
All this sounds very progressive, compared to a past situation where operators often had to snap up whatever band was available at any one time, even if it was not optimal. But we are talking about government agencies here. Putting two auctions together doubles the complexity, the bureaucracy and the opportunity for delay. So while countries like Sweden, sticking to sequential auctions, already have 2.6 GHz out of the door, and are hurtling towards 800 MHz sales, cellcos in the UK and India are living with protracted delays in being able to acquire any spectrum at all for the next generation of their services.
This is leading to calls for auctions to be decoupled again, especially in the interests of new entrants with new service models. In the UK, existing cellcos can make do with their existing holdings for some years, now that they are almost certain to be able to refarm their GSM bands for 3G (though the possible redistribution of 900 MHz, which only Vodafone and O2 hold, remains a sticking point even for that). But operators hoping to acquire a license to create a new model and introduce added competition and new types of services are increasingly frustrated. The UK was supposed to be the first European country to auction 2.6 GHz licenses, back in 2008, and that would have provided a favourable environment for a new mobile player (or a returning one like British Telecom) to leverage a new technology like WiMAX, to leap ahead of the existing 3G incumbents. The longer the delay, the less attractive that opportunity is, because the new players will be running head-to-head with the incumbents’ LTE plans, as that technology becomes available.
In India, even entrenched mobile operators cannot afford to wait much longer for new spectrum, since exploding demand is overloading their current networks to breaking point. However, there is still a case for separating the two auctions, since the problems that are delaying the 3G sale (mainly disputes over when the Department of Defense will vacate spectrum, and the reserve prices) do not affect the 2.5 GHz band, and at least the WiMAX providers could make a start on addressing the country’s glaring shortage of broadband services.
Yet the two sales remain bound together, and delayed yet again, probably until the new fiscal year starting in April, according to government sources. The auctions have been delayed many times, most recently until January 14 and then February 12. This will be bad news on all fronts. The wireless infrastructure suppliers may not be expecting quite a big a boost as from Chinese 3G, given the price competition in India, but the sheer size of the market is still vital to a still pressurized sector; India is probably the world’s largest potential market for WiMAX; and the Indian cellcos themselves badly need the new spectrum to add desperately needed capacity and launch higher margin services.
All this is creating a groundswell of lobbying for the WiMAX sale to go ahead regardless of the constantly changing rules on the number, price and conditions of the 3G licenses. An internal note authored by two units of the Department of Telecom, and quoted in the Economic Times newspaper, bemoans an “amazing array of inconsistencies in the government’s approach to 3G mobile licenses, which are unlikely to be granted anytime soon”. As the paper points out, further delays could “stymie the progress of the telecoms sector, the vaunted success story of liberalized Indian infrastructure. This hurts the economy, forgoing enormous, life changing growth potential in rural India.”
It voiced the views of many elements of the telecoms industry, saying: “The time has come for the government to delink the auction of licenses for wireless broadband services using WiMAX from auction of 3G licenses. The government should move ahead with WiMAX spectrum auctions, without waiting to complete 3G auctions. This would open up much needed new revenue streams for the industry. The issues that hold up licenses in 3G do not apply to WiMAX. The government should move ahead on this front.”
The most likely opposition to this view would come from LTE suppliers, reluctant to see WiMAX get such a powerful beach head in the world’s second largest market. Such parties would be pleased to see 2.5 GHz allocations delayed, especially as the Indian DoT and regulator TRAI said last week they were considering further auctions “very soon” after the 3G and WiMAX transactions, to support ‘4G’ services. The first would be digital dividend spectrum around 700 MHz, which India has long planned to bring into play ahead of most of the world. As 2.5 GHz will be used almost entirely for WiMAX, because of the immediate availability of the technology, this leaves 700 MHz as the only near term band with international support for LTE suppliers to target, although HSPA and WiMAX may be more likely options. LTE proponents like Ericsson and Qualcomm have already been accused by the WiMAX community of lobbying to delay mobile broadband auctions in countries like Brazil, until their favored technology was ready.
This was also a factor, as we saw above, in a very different mobile economy, the UK, where timely spectrum availability could have given WiMAX a foot in a firmly GSM-focused door. That door may be closing, as the country is now likely to wait until 2011 for its 2.6 GHz sale, leaving carrier plans for LTE or WiMAX on hold. Kip Meek, the government appointed independent spectrum broker, says a 2010 auction is now “a real stretch” and 2011 is the probable date.
The auction was originally delayed by a legal challenge from the UK arms of T-Mobile and O2, which argued that they could not put a realistic valuation on the new bands until regulator Ofcom had clarified its policy on refarming GSM spectrum. That in turn would require a settlement of the long running dispute over whether O2 and Vodafone, the only holders of that lower band GSM spectrum in 900 MHz, should be forced to redistribute their holdings, which are valuable for cost effective rural coverage.
Now, other factors are causing delays - the proposed merger of T-Mobile UK and Orange UK, which could break existing spectrum caps even without new allocations; and the decision to sell the 2.6 GHz licenses in tandem with the digital dividend spectrum around 800 MHz.
Meek’s latest update was reported by online newspaper ZDNet UK (www.zdnet.co.uk), at a Westminster eForum on the spectrum implications of the digital switchover. Meek said the outcome of his proposals for 2.6 GHz allocation was “uncertain” because of operator concern over the TMo-Orange merger, and there is threatened legal action from British Telecom. He told ZDNet: “The government will be considering its response to the consultation paper, [but] that process is taking longer than originally anticipated. They will have to take a view whether, in the dying months of this government, they will want to push through with this particular set of proposals.”
Meek wants Vodafone and O2 to receive 800 MHz allocations only if they surrender an equal amount of 900 MHz spectrum. And TMo and Orange should be able to bid only for limited amounts of 2.6 GHz, unless they give up some holdings at 1.8 GHz or 2.1 GHz. He has also proposed making 2.1 GHz 3G licenses indefinite in length, in return for commitments from cellcos to improve 3G coverage. Currently, they are due to expire in 2020.
This last idea is the one that could incur BT legal challenges. The incumbent has described it as “a gift of several billion pounds from the UK taxpayer to the mobile operators”. BT spun off its own cellular arm, which became O2 and is now owned by Telefonica. BT was widely expected to bid for a 2.6 GHz license and run WiMAX in it, primarily as a wholesale network or to support new business models such as machine-to-machine.
Après maintenant, le déluge
By Andrew Mitchell
By Carl Howe, Director, Yankee Group Research
That’s French for “After now, the deluge”. And that’s about how I feel about today’s CES opening in just 2.5 hours. It’s going to be a very busy day.
But before the networks melt (or at least glow bright red with heat) from 120,000 people blogging away, I thought I’d recap a few highlights of the last 18 hours deserve some note. Instead of running down the facts of each announcement, I’ll let links to New York Times articles do that for me; I’ll just add my reasons why the events are significant. Specifically:
- Vendors are pushing 3D as the next high-definition. Every flat panel vendor here is hoping that 3D televisions will be the next must-have upgrade. I must admit that I’ve been impressed by the demos here; I saw Sony’s 3D technology with Taylor Swift performing last night, and it was one of the best broadcasts I’ve seen with passive glasses. All that said, with the inconvenience of 3D glass and most flat panels only one to three years old in the U.S., I don’t think most consumers will even consider replacing them for another 3 to 10 years, dooming the trend until the latter half of our new decade. I’ll be touring Panasonic’s booth this morning (it’s the size of a football field), and may have to say about this tonight.
- TVs are adding connected apps in a big way. Both Panasonic and Sony are boasting Skype conferencing on their connected TVs, and that’s far from the last app they’ll be adding. Whether built-in or add-on connected TV apps are the wave of the future, there’s no question in my mind that connected TVs will be a bigger trend than 3D in the short term, and it may be true later in the decade as well. Vince Vittore’s report published just yesterday is a timely analysis of this trend.
- Consumers may be pushing companies to become more vertically integrated. Apple’s been able to win over consumers over the last decade because it can orchestrate hardware, software, and media to create a single unique consumer experience. Microsoft’s one success story in the last decade has been XBox, which is similarly vertically integrated. Sony’s recent successes have been with Playstation Network melded to its PS3 and PSP gaming systems. Despite the appeal of open partner, these vertically integrating offerings are winning consumers and creating profits. The very fact that Google had to create its own Android phone to drive adoption suggests that we may be seeing the open pendulum swing back toward soup-to-nuts offerings.
- Cars are moving to glass cockpits. Despite the dangers of distracted driving, car companies are madly rushing to put Internet connected screens in car dashboards. I don’t actually believe that Anywhere connectivity is the juice that makes this trend take off, but rather the replacement of traditional car instruments with electronic icons, maps, and controls. By the end of this decade, analog guages in cars like we do in airplanes today: amusing antiques.
- Tablets just aren’t here yet. Despite the excitement over Apple’s rumored iTablet and Steve Balmer’s announcement of an HP tablet last night, the reality is that these connected devices are still a dream rather than reality. The litmus test? Try to buy one. It will take more than CES to get this trend moving, but it will probably take Steve Jobs bringing tablets down from his mountain later this month to make it reality instead of vapor, and even then, no one’s going to be using those tablets in any volume until at least Q2.
Ericsson and Huawei bicker over Nordic region’s LTE headstart
By Caroline Gabriel
Late last year, TeliaSonera won the laurels that most thought would go to Verizon or DoCoMo, of the first live LTE network, admittedly a small one, covering the downtown area of Swedish capital Stockholm. Although Telia is working with both Ericsson and Huawei on its trials, this system runs on the local vendor’s kit, but shortly afterwards a row broke out between the suppliers over their various Scandinavian networks.
TeliaSonera customers are being issued with Samsung LTE dongles to test the network. Kenneth Karlberg, president and head of mobility services at Telia, said: “Thanks to the successful cooperation with Ericsson we can offer 4G to our customers in Stockholm earlier than originally planned.” A commercial launch of the trial network in Oslo, Norway, supplied by Huawei, will follow soon.
The carrier plans to expand its LTE reach to Sweden’s 25 largest cities and vacation areas, and Norway’s four largest cities, in 2010. The capital expenditure for this will amount to SEK500m ($70m). It will also kick off a pilot LTE network in Finland in the first quarter, in its 2.6 GHz spectrum acquired in the recent auction. All three of these Baltic countries have been ahead of the curve in selling 2.6 GHz licenses, enabling the early shift to commercial 4G services, in contrast to the hesitant process towards opening up new frequencies in the major western European economies.
In parallel with the LTE roll-out, TeliaSonera also promises “extensive expansion of our Turbo 3G [HSPA] networks to increase coverage and capacity”. Like most 3G carriers looking to introduce LTE to their networks at an early stage, Telia will go for an overlay approach, with a common IP core integrating 2G, 3G and 4G systems.
It stressed that it has not confirmed which suppliers will provide the networks for the national roll-outs. “Evaluation of suppliers for TeliaSonera’s common 4G core network and radio networks in the Nordic and Baltic countries is in progress and vendors will be selected in the beginning of 2010,” said TeliaSonera’s head of mobility services, Kenneth Karlberg. “Choosing Ericsson and Huawei was a result of what vendors were available at the time for these two cities. The next deal is for a bigger roll-out… and now there are more vendors that can provide LTE equipment - the evaluation should be completed in January.”
To attract users to the Oslo and Stockholm networks, Telia is offering ultralow introductory pricing. In Stockholm, consumers signing up for a 12-month contract pay just SEK4 ($0.56) a month until July 1 2010. They receive the Samsung dongle free, with a free upgrade to Samsung’s multimode 3G/LTE dongle when it is available in the second quarter next year. From July, customers will pay SEK599 ($83) per month for 30 Gb of data, and the modem will be charged as an extra, as yet unspecified, fee. In Oslo, the same procedure applies, except the initial monthly cost until April 1 2010 is NOK1 ($0.17), after which time the monthly tariff is increased to NOK699 ($120).
Mikael Bäckström, Ericsson’s president for the Nordic and Baltic regions, told Total Telecom : “In the Nordic market in general there is a big uptick in users and revenues when it comes to mobile broadband. It is a competitive market also, so being able to offer more bandwidth, more capacity and higher speeds than your competitors is important.” He claims customers will experience speeds of 20 Mbps to 80 Mbps and this will increase as the network is optimized, adding that LTE will enable new revenue streams from enterprise services, and will “enable governments to take services that were only accessible using the fixed line internet, and make them available on the mobile, and offer a more advanced service.”
Huawei may not have grabbed the kudos of providing the first live LTE system, but it quickly claimed that its Oslo trial network had reached peak download speeds of 96 Mbps, compared to 44 Mbps on the Ericsson system in Stockholm. This bit of one-upmanship becomes rather less interesting when you realize that the Oslo network uses 20 MHz of spectrum - the channel size usually assumed when LTE performance statistics are quoted, but in reality rarely available to cellcos. By contrast, the Stockholm build-out has only the more common 10 MHz. The Oslo network goes live this month.
Huawei has also won another LTE deal right in Ericsson’s home country, with Sweden’s Net4Mobility, a joint venture between Nordic carriers Tele2 and Telenor. It aims to start rolling out LTE in 2010 and cover 99% of the Swedish population before the end of 2013, starting with densely populated areas. The deployment also includes a provision to increase the number of 2G base stations for voice traffic by 30% to 50%, to improve coverage indoors and in rural areas (and allow the operator to put off the voice-over-LTE decision for a while).
Unusually, Ericsson issued a statement about its failure to get the contract in its homeland, and was keen to suggest that Huawei had won on price rather than superior performance. “We are of course disappointed that we did not manage to reach an agreement with Net4Mobility … We would very much liked to have delivered this LTE network in our home market. In the negotiation process we went as low as we could in terms of price but it was not enough,” said the firm.
A 4G Season for Reflection
By Andrew Mitchell
For the 4G wireless industry, 2009 has been a year of unprecedented challenge and opportunity. The economic downturn of 2009 has been felt across almost every industry sector and in just about every country on the globe. Despite tough times, the wireless industry as a whole continued to hold onto a momentum of innovation and growth in 2009.
While talk of LTE began to dominate industry headlines, WiMAX technology continued to deliver a healthy roll-out of deployments in emerging markets as well as in the rural markets in the US. CLEAR 4G WiMAX service launched in Portland, Oregon back in January and has completed deployments to cover 28 other US cities. The first round of the R-USA BTOP funding generated over 2200 applications totaling over $28 billion in requests. WiMAX technology has become an integral component of many of the solutions represented by successful applications.
The challenge of gaining access to 4G spectrum was seen in many countries throughout the world and probably most noteable in 2009 has been the complex and often controversial auction in India. Long awaited access to 700 MHz spectrum in the US was finally granted after the overdue completion of the US’s DTV initiative. While there has been some headway made in the ongoing battle for access to 4G spectrum, it remains very clear that regulators in both emerging and mature markets will need to step up their efforts in 2010 and beyond to balance the interests of incumbent operators and game-changing start-ups.
It was an interesting year for LTE technology. The expectations and stakes are high as Ericsson will move ahead with an LTE strategy as a result of their successful bid for the CDMA and LTE assets of Nortel. The bidding process was intense, fraught with controversy and difficult to call, with RIM and Nokia Siemens Networks eventually losing out. Ericsson’s imminent opportunity to advance a return on their investment will come from Verizon’s announcement of Ericsson and Alcatel-Lucent as vendors selected for its 2010 4G network roll-out.
The road to a more pervasive mobile Internet still has some bumps and curves ahead. While advances in performance, power consumption and flexible radio technology delivered powerful new platforms like Samsung’s Mondi, 2009 didn’t deliver a 4G smartphone on the CLEAR network. Clearwire’s CEO, Bill Morrow promises that one will be available for the 2010 holiday season though. Some mobile applications developers made headway in 2009, distributing their wares through Apple’s App Store for example, but not everyone in the ecosystem is pleased with the current model of application distribution. Developers cite reasons for their frustrations - too many platforms to support, too many carrier imposed restrictions and too little return for the effort to overcome these obstacles. A continued move toward an open mobile environment will need to remain a priority for 2010 to drive the pervasiveness that the industry and consumers need.
In spite of the rough economy, the wireless ecosystem demonstrated a commitment to the importance of industry events. Key conferences such as 4G World, CTIA Wireless 2009 and Mobile World Congress attracted innovators and thought leaders focused on creating new technologies and business models that will shape future economies and improve quality of life for all. All of the 4G Trends editorial staff have enjoyed participating at these events and following the 4G industry and are looking forward to an exciting and productive 2010.
Until our next issue on January 6, 2010, we wish our readers and friends a safe and happy holiday season and a peace filled and prosperous New Year.
Uplink performance could be LTE’s dark horse advantage
By Caroline Gabriel
In the old days of the Internet, it was all about the downlink, with most activities centered on browsing and downloading. The slower performance of wireless technologies on the uplink did not matter, but increasingly the focus is shifting, as two-way applications like social networking become key drivers of mobile web behavior. As LTE looms on the horizon, then, carriers are interested in its uplink behaviour for the first time, and several companies are trying to enhance its credentials.
In the UK’s ‘Silicon Glen’, Cambridge Consultants is claiming a breakthrough in improving LTE uplink performance, without any changes needed to handsets. This is done with its Duel (Dual-Domain Uplink Equalizer for LTE) receiver technology, an alternative to the conventional receiver design, MMSE. Duel is implemented as an extension to LTE base station receiver designs, providing a set of “unique and highly complex algorithms” that generally run on DSPs, but can also be deployed on FPGAs or ASICs. No changes at the device end are required to gain the performance benefits.
Uplink performance is increasingly important as mobile Internet usage patterns shift from downloading of web information to two-way communication and content generation. But as with all mobile platforms, LTE uplink remains significantly slower than downlink and the high data transfer rates promised by LTE require ideal channel conditions, such as low mobility or high diversity.
Duel, says Cambridge Consultants, improves uplink performance in all channel conditions, by better exploiting the inherent advantages of LTE’s uplink technology, SC-FDMA. It has developed a test platform that includes a conventional MMSE receiver alongside the Duel device to allow their performance to be compared on the same received signal. The test system uses a combination of devices from Cambridge Consultants’ proprietary Viper development platform and from partner picoChip.
The SC-FDMA technology is the most novel aspect of the LTE standard, and the most important area where it differs significantly from fellow 4G platform WiMAX. In many ways, the two systems are technically very similar, and both use OFDMA on the downlink, but while WiMAX also uses this technology for upload, LTE combines two approaches. LTE proponents believe SC-FDMA is more efficient for fast uploads and good coverage because of its multipath qualities, but also that it conserves battery power better than OFDMA. It allows devices to transmit low power signals without the need for expensive power amplifiers. SC-FDMA resembles OFDMA in many ways, but it splits and recombines uplink transmissions on a single channel, which is a low power approach that could be important in counteracting the otherwise power hungry nature of early 4G silicon. It uses a Fourier transform to convert OFDMA’s separate subcarriers on separate frequencies into a different, time-domain, form, which could improve its operation at high mobility.
While companies like Beceem are working on improvements to WiMAX’ uplink too - the chip designer released its 4G Turbo, with enhanced uplink reach and performance a key feature - the next WiMAX standard, 802.16m, is expected to use SC-FDMA for its uplink modulation too, to take advantage of its reduced peak-to-average power ratio, which is high in OFDM technologies.
Do Your Favorite Places Offer QR Codes?
By Andrew Mitchell
By Jon Paisner, Senior Analyst, Yankee Group
Dec 7th, Google announced its next step in the journey towards enabling an Anywhere lifestyle for both businesses and consumers. Although trials have occurred over the past year, Google is now launching a national campaign by sending out window decals which contain QR codes to over 100,000 local businesses as part of the Favorite Places program. By scanning these QR codes, consumers will be given detailed directions through Google Maps, read detailed reviews of the merchant and possibly even receive a mobile coupon.
Following upon the recent successes of integrating mobile into the Black Friday shopping experience (see blog post by fellow Yankee Group Analyst Dimitriy Mochanov titled “Retailers Beware! The Mobile Shopper is Here!”), proliferation of smartphones and data packages put these capabilities in the hands of the consumer. With this Google Favorite Places announcement, local merchants do not need to invest in a mobile platform to provide reviews or drive foot traffic into the store. Merchants leverage Google’s investment and do not require infrastructure to drive consumers to their store. However, once merchants offer mobile coupons they will need 2D barcode scanning capabilities at the point of sale (PoS) to redeem the coupon from the handset. Now that we’ve gotten the details out of the way, let’s examine what QR codes may mean for general consumers and merchants.
QR codes can provide value but how much value depends on how merchants use them:
- Directions through Google Maps: In this first iteration, Google is providing a consumer with step by step directions to a store. The problem is that the consumer is scanning this QR code once they are already in front of the store. The value to the consumer here is limited. However by integrating the same QR code within your online advertising? Now Google Maps can provide walking and/or driving directions which becomes a real value add to the consumer.
- Ratings & Reviews: Providing reviews for a restaurant or store is limited when the consumer is right outside. Why not just see if the restaurant is crowded or try to visually see the presentation of the food yourself? However the value here for the QR code is to let the consumer read reviews of the level of service provided and intangibles such as cleanliness that the consumer cannot easily observe from outside the establishment.
- Mobile coupons: A consumer is outside your store and has scanned your QR code, but how do you get them to come inside? You have a captive consumer ready to interact so the mobile coupon is the best option to get them inside. The other options discussed above provide information while mobile coupons provide transactions and revenue. Utilize them.
After scanning QR codes this morning on my 2 year old Blackberry Curve, here are my results (which are unscientific but at least show what an average consumer will be up against):
- Size matters: When scanning the QR codes directly from my computer monitor, the size of the code matters. None of the Google Favorite Places in Boston worked with 3 different app readers on my Blackberry Curve. Once I enlarged the size of the code they worked just fine. It appears that the stickers Google is issuing for merchant windows are large enough where this doesn’t become a problem but is still worrisome for merchants looking to integrate mobile into their offline and online marketing strategy.
- Consumers don’t discriminate, neither should technology: With no common standards utilized throughout the US, app readers work on different types of codes. Consumers don’t discriminate between QR codes, Datamatrix codes, or any other proprietary code but the app readers themselves are not universal for all of the various code types. In the mind of the consumer they think “hey, here’s a funky looking code, I’m sure my app reader will work” and not understand the nuances between the technology. If you don’t believe me, read the reviews in the various application stores. The 3rd party app readers have very poor ratings but it is often not a technology problem reading the codes. Consumers are just trying to read interoperable code types.
- Cameras aren’t a limiting factor: Using a 2 year old Blackberry Curve and an original iPhone (without autofocus) to test out various codes, these older camera’s worked just fine. New phones with high resolution cameras and autofocus should experience even lower failure rates.
A whole host of other issues still combat this industry trying to gain popularity anywhere outside of Japan. For QR codes to become popular inside the US very soon merchants will need to deliver value to consumers, whether that is information based or through cost savings. In the end, the merchants are looking to drive transactions and revenue and the best opportunity to do so is with mobile coupons. These coupons will drive price sensitive consumers to open the door and come inside.
LTE will create more silicon shake-up, and chances for WiMAX players
By Caroline Gabriel
The landscape for wireless device chipmakers is changing rapidly. The baseband suppliers have reduced in number, driven by the concentration of handset power in the hands of a few giants. Qualcomm, Infineon and ST-Ericsson are the main survivors, with Broadcom tapping at the door and MediaTek threatening to break out of its low cost Chinese base.
But as phones get more complex, and wireless becomes embedded in a host of other rich devices from smartbooks to e-readers to media players, there are new opportunities for silicon companies to attack the slots around the baseband. So the mobile applications processor market is hotly contested, with Samsung, Nvidia and others vying with the dominant Texas Instruments and the baseband/SoC players - and of course, Intel hovers in the wings too, eyeing the smartphone for x86/Atom, by way of the safer territory of the netbook. Samsung is planning to spend around KRW7 trillion (about $6 billion) on expanding its semiconductor business in 2010, according to a Korea Herald report - most of that may be going on the key DRAM and NAND flash memory lines, but some will certainly boost the mobile processor activities.
At the LTE stage, these shifting sands will result in a more complex web of alliances to address a diversified set of form factors for mobile broadband devices. Qualcomm will still aim to provide everything a device maker could require, with multiple CPUs and graphics processors (GPUs), combined with a widening range of optimized software. But many others will form partnerships to strengthen their hands - perhaps in graphics (especially as more non-graphics functions are offloaded to the GPU), as seen in Intel’s increasing stake in processor IP firm Imagination Technologies; perhaps in LTE, the great hope for several chipmakers that have cut their OFDMA teeth in WiMAX, but know they will need powerful friends to play in LTE, where the market will be dominated by bigger fish.
The WiMAX device silicon business has been very fragmented, and several of the players have recently announced parallel WiMAX/LTE roadmaps, harnessing their experience in the older, but very similar, technology. Sequans, WaveSat, Altair, GCT and Beceem are among those with an LTE plan, though they are all at very different stages in terms of actual products. Altair looks advanced, having a baseband and RF transceiver for nearly all the frequencies and modes supported by the currently fragmented LTE segment. This is now included in a PC card reference platform that can be used for field testing. Eran Eshed, co-founder and head of marketing and business development, does not believe there will be room in the LTE device ecosystem for a new chip start-up, but that a couple of WiMAX players could make it, probably through an acquisition or by providing early stage LTE weapons to a larger baseband provider that would otherwise fall behind the frontrunners, Qualcomm and STE.
He, of course, believes Altair can successfully bridge the WiMAX-to-LTE divide, and it has a common, software defined platform that supports both standards, as well as OFDMA-based XGP (see separate item). The need for operators, and therefore device makers, to integrate legacy and new networks gives smaller chipmakers their opportunity he argues, especially when the strong players in the existing platforms fail, for one reason or another, to develop an LTE offering at an early date. A particularly low hanging fruit might be a partnership with one of the five Chinese silicon vendors that make basebands for TD-SCDMA, none of which, he says, has a viable solution for China Mobile’s migration to TD-LTE. This will be one clear option for an Altair alliance, or for Sequans, which recently gained funding from Motorola and Alcatel-Lucent, largely geared to those firms’ bids to provide TD-LTE kit for China’s proposed network at Shanghai Expo in 2010 (both are working with the French firm, while Huawei has silicon from its own Hi Silicon subsidiary). Altair, which is also participating in those trials, also does not rule out a strategy to ‘help out’ more broadly based baseband/processor makers, such as MediaTek or Marvell (which also have TD-SCDMA, but not TD-LTE) or even Broadcom or Infineon, which have not shown their hand on any substantial LTE assets as yet. Knowledge of the very different ways that OFDM technology works, compared to 3G, could be a valuable short cut for any of these firms in a difficult transition to a new radio world. Eshed also believes that “companies without RF transceiver products will find it extremely difficult to compete in LTE. Such companies could make it in WiMAX as they used RFICs from pure play companies such as Maxim or PMC Sierra. The situation in LTE is completely different, as such RFIC companies understand that this is a game of the large cellular chip companies and will not offer such products to the market.”
Infineon’s likelihood of wanting help from a WiMAX expert may have receded, given a new strategic alliance with Nokia to collaborate on silicon for the next iteration of the LTE standard, LTE Advanced, which will find its way into mainstream devices 8-10 years from now. The agreement highlights the new balance of the handset chip industry. While the biggest manufacturers are increasingly cutting back on developing their own customized designs, in favor of buying off the shelf (hence the radical change in Nokia’s relationship with its former near-monopoly supplier TI), they actually want increased control and differentiation in the emerging technologies. So we have seen Apple acquiring PA Risc and upping its stake in Imagination, possibly with a view to designing its own processor architecture for future iPhones, relegating Samsung from key processor provider to manufacturing partner. And of course, Nokia shook up the whole sector in 2007, and forced TI’s virtual exit from basebands, when it decided to introduce a second source for each category (Infineon for GSM, Broadcom for EDGE, STMicro for 3G). Last year, it extended Broadcom’s reach into low cost 3G device platforms, and added Qualcomm for north America and Intel for the netbooks.
Now it has enhanced Infineon’s role, and the German firm is working with Nokia on LTE Advanced in a strategic development role, of a kind TI used to command, and which the Finnish giant also has, in HSPA+, with ST-Ericsson. The LTE Advanced project could pave the way for extended cooperation in other areas, and possibly a place for Infineon in high end Nokia devices.
The new partnership focuses on developing handset platforms that could deliver gigabit data transfer rates, as required by ‘true 4G’ standards such as WiMAX 2 and LTE Advanced. The collaboration is non-exclusive but Nokia’s first announced R&D in this area, and will center on developing advanced RF transceivers, based on Infineon’s RFICs and closely integrated with Nokia basebands.
The resulting designs could go beyond giving Nokia a headstart in the new generation of LTE after mid-decade, and the work could also be fed into complete modem solutions for HSPA+ and basic LTE. Nokia is increasingly prepared to license out its modem designs for revenue and broad industry influence and will create reference platforms based on its Infineon project.
Infineon did not say how many engineers would work on the project, or what the deadlines are, but indicated that the results would be contributed to the LTE Advanced standards process. “We are grateful to expand our successful collaboration with Nokia beyond our current platform and RF activities,” said Professor Hermann Eul, member of the Infineon management board, in a statement.
“Taking advantage of each company’s expertise as leaders in their respective fields, this cooperation will help to deliver standard-based, industry leading solutions for mobile internet devices,” said Pekka Sarlund, VP of wireless modems at Nokia.
While Infineon is far from out of the woods of its recent financial turmoil, its strongest business has been in its wireless activities, and it has been moving up the smartphone food chain, most famously by getting into the iPhone. The German firm has also scored with its low power, single-chip GSM architecture, as used by Nokia. It first entered the mobile RFIC space 10 years ago shortly after the spin-out from Siemens.
