Your browser is not supported. Please update it.

28 May 2007

WiMAX World: Forum gloomy about EU, but elsewhere 802.16e can still beat LTE

The ripples from Ericsson’s recent decision to stay out of the WiMAX market were certainly visible on the river Danube as Austrian capital Vienna hosted the second WiMAX World Europe conference this week. Two questions often asked about mobile WiMAX were voiced more strongly than ever in the wake of the Swedish giant’s rejection – is there really an opportunity for 802.16e in western Europe, the heartland of GSM/UMTS; and is 802.16e in danger of losing its window of leadership over the main alternative mobile broadband offering, LTE? Mo Shakouri of Alvarion, the public face of the WiMAX Forum, said in a keynote address that he thought western Europe was closed to large 802.16e deployments, a gloomy outlook not helped by the decision by UK regulator Ofcom to delay its auction of 2.5GHz licenses until 2008, thus staying the hand of the greatest hope for WiMAX among the European majors, British Telecom. And the message from service providers on the podium and in the audience was clear – no more delays and changes to the standards, certified equipment needs to roll out on schedule now or risk losing the game to LTE.

Operators were horrified to hear some serious parties airing the view that the industry should focus on leapfrogging LTE by concentrating its efforts on the next iteration of the WiMAX standard, 802.16m (not yet finalized). It is now ancient history how the early focus on 802.16d pushed back the more important 802.16e standard, which with the advent of Samsung and WiBro became incompatible with its fixed predecessor. But now there has been another factor pushing the timescales for mass market equipment out – the tendency for large vendors to ignore the ‘wave one’ profile of 802.16e and wait for the more powerful wave two. Wave two will certainly support better business cases than wave one, with its mandating of MIMO and other techniques for improving efficiency and capacity, but there is also a limit to how many times the industry can delay putting real products into the field. Rather like the perfectionist writer refining a novel over and over again until the publisher demands the advance back, the WiMAX suppliers are in danger of creating a beautiful network for which there is no longer any shelf space reserved – and talk of waiting for ‘wave three’ or even 16m send shivers down the spine of the operators that are relying on making an early move into mobile internet.

Even more perturbed are the operators in territories that are far stronger for WiMAX than western Europe – emerging economies where wireless is the most available connection, where 3G has often not taken hold yet, and where infrastructure is not keeping pace with growth. In most cases, as a keynote speaker from Pakistan’s Mobilink pointed out, these providers want to offer fixed broadband access to the home, a basic model that requires volume economics but not necessarily sophisticated handsets or vehicular mobility. These providers want to see prices coming down and to get their hands on certified equipment, even if it falls short of the maximum functionality of the 802.16e – and then think about mobility in years to come.

Such concerns about WiMAX will gladden the hearts of the UMTS community, which is heavily focused on LTE. Currently, that standard is unfinalized, but the industry has certainly responded to wake-up calls from operators, like that from Vodafone CEO Arun Sarin at the 3GSM show in February, urging vendors to accelerate development of LTE to give 3G carriers a growth path – or risk, he implied, losing the cellcos to WiMAX. Ericsson’s single-minded commitment to LTE has helped this effort rise to the challenge of narrowing the gap with 802.16e – removing the only real factor, availability, that would make WiMAX a better choice than LTE for a provider with existing investment in UMTS (although LTE will be very different from current 3G, the history and R&D focus of its major supporters, and the input of the UMTS cellcos themselves, should ensure that it will be a somewhat less painful migration path from 3G than other options). LTE will, of course, meet its own delays and obstacles along the road, but it will benefit from its major overlap with WiMAX (up to 80% of R&D is in common, according to Motorola), allowing for various shortcuts in the development roadmap.

Large cellcos are already testing 802.16e and pre-standard LTE prototypes in parallel, which appears to justify the decision by companies like Motorola and Nortel to carry out parallel development of both platforms, reusing as many elements as possible. Having all but exited UMTS, these two will need to demonstrate real advances in OFDMA-based systems to have a good chance of penetrating the LTE market, at least among traditional cellcos. They argue – and perhaps surprisingly, some UMTS cellcos are agreeing – that Ericsson has put itself at a huge disadvantage by choosing a non-WiMAX program, because it will have to move from CDMA-based to OFDM-based architectures in one leap, rather than having WiMAX expertise to prove its case in LTE.

Such comments raise the question of whether some vendors will use 802.16e primarily for that purpose – to give themselves early experience and credibility in the OFDMA/MIMO architectures that will underpin mobile broadband, but with a view to putting LTE at the center of the strategy once it becomes a reality. This is unlikely to be true of Motorola, which has both the urgent need and the scale to back WiMAX uncompromisingly and therefore put itself in the driving seat of that technology, as Ericsson has in GSM and UMTS. But there may be more doubts around the other UMTS stalwarts, Nokia Siemens and Alcatel Lucent, both of which will aim to defend their traditional customer bases among the large cellcos by making early moves to LTE – and could, especially if they have not gained major market traction in WiMAX, then defocus on 802.16e.

All these are risk factors for mobile WiMAX, which are inevitably highlighted most harshly when in western Europe, the most unforgiving market for any technology that has not been spawned by the GSM community and the cellular business model. But 802.16e’s weaknesses in the UMTS lands are its strengths where 3G has not yet been built out – in particular, its heritage in the PC internet business model. For an operator with the need to launch services quickly, but without a short term focus on mobility or voice, HSPA remains too voice- and downlink-focused and inadequate to replace DSL indoors; LTE is too far away and too driven by the requirements of the large mobile carriers; and so WiMAX appears a strong choice for deployment in 2008-9. Above all, it comes with a readymade ecosystem – its use, PC industry-style, of off-the-shelf components (witness the early commitment to the OBSAI base station interface standards) and existing, open software platforms means that, once the base stations and devices actually ship, there will be far less work to be done by integrators or operators on creating the rest of the system around the infrastructure. The prompt entry of the Taiwanese ODM players, and the early signs of a parallel ODM sector growing around WiMAX in India, indicate how rapidly this technology has evolved to support consumer electronics economics and highly integrated systems. Intel may have focused the debate for too long on high end, power hungry CPE such as laptop cards, but in Vienna it was talking up its low power advances and responding to some of the core requirements of the emerging markets base. And undoubtedly, the PC chip giant has imported the openness and economics of its traditional market to the wireless world, and that is looking to be its greatest contribution -aside from its marketing machine – to the 802.16 cause. WiMAX in western Europe is likely to remain confined to niche business models in 3.5GHz spectrum, but elsewhere its opportunities remain impressive – just as long as the certification and product shipment process suffers no further delays.