WiMAX specialists can benefit from 16e delays, if they can survive the wait
Published: Monday 29 January, 2007
It's all starting to sound horribly familiar. Delays to the appearance of fully functional WiMAX standard products, forcing early adopters to choose between two risks - go with pre-certified, semi-proprietary kit with all the functions they want; or select first phased certified equipment that is only suitable for limited applications. The scenario is familiar from 2005, the year when operators hoped 802.16d fixed WiMAX equipment would be certified, with price points falling and functionality rising. In fact, there was more than a year to wait, and though prices did fall, this was largely down to increased competition and the exceptionally early stage entry of the Taiwanese ODMs, rather than standardization. As for functionality, as the gap between 802.16d and the start of 802.16e certification appeared to narrow, there was less incentive for vendors to put significant development effort into the former product, and R&D budgets were diverted to taking an early position in 'e', preferably to forestall the tier one manufacturers, most of which were skipping 'd' altogether.
Now the same dilemmas are afflicting 802.16e adopters, but this time they are far more critical, both for operators and vendors, because such grand financial plans have been vested in a technology that is promising mobility, a wide variety of device types including handsets, and the functionality to support wide area, carrier class broadband services. It seems that the timescales for 802.16e certification are being pushed back, according to the WiMAX community gathered en masse at the recent WCA conference in California, and operators have little chance of seeing products with the all-important wave 2 certification this year - wave 2 takes 802.16e from a fairly basic platform to one that promises to be suited to personal broadband and nomadic, plug and play operation, even if there are still many kinks to iron out to achieve full wide area mobility.
These delays will be most serious for vendors that are expecting serious revenue from full scale 802.16e deployments by the end of 2006. Most large operators will not have expected to implement more than trials this year anyway, though the persistent delays in the certification that kicks off the hoped-for downward price pressure will be a source of nervousness. For other parties, there will be opportunities in delay. Just as some proprietary BWA vendors welcomed the postponement of 802.16d and the start of the move to commoditization, taking advantage of the new interest in BWA that WiMAX had engendered to sell greater volumes of semi-proprietary, 'WiMAX-ready' kit, so the specialists in 16d and in pre-certified 16e will see a longer period than they had expected to push their solutions at a higher margin.
The main players in pre-certified 16e platforms are Samsung and Navini, which have real products on the open market, while others are just producing sample equipment, often for specific trials. Samsung has the advantage that the wave one 16e certification was all but modelled on its Wi-Bro technology, now rechristened Mobile WiMAX to emphasize the claimed easy route to wave 2, considered by most to be the 'real' 16e. Navini has a more tricky problem. Along with Samsung, it has been very early in achieving the highly complex task of creating fully nomadic, plug and play CPE, such as laptop cards, which necessitates the incorporation of the advanced antenna techniques of 16e wave two (notably AAS and MIMO) - without which the dream of non-line of sight, indoor, portable devices remains a dream. While Navini can exploit its headstart in this important product category, it also faces the start-up's dilemma - how far it can persuade the market that its Ripwave MX system, which also features beamforming and claims to be ready for wave two certification, is a safe bet, thus generating sufficient revenue to justify the high cost of developing these mobile-capable products, and enabling the company to regard the hiatus period before wave 2 testing as a bonus (longer period of limited competition and higher margins) rather than a threat (longer period to survive without the volumes that a fully standardized market can bring).
The 16d vendors have something of the same problem, although their claims that their products will offer an upgrade to 16e have gained considerable credibility since the launch of Intel's Rosedale 2 chipset, which straddles both the usually incompatible WiMAX standards. Some of them face a problem that is the direct opposite of Navini's - while Navini has taken the risk of committing from day one to a complex and expensive R&D route, in order to take an early position in fully plug and play personal broadband, some 16d vendors have scaled back their efforts to support only fairly simple platforms, confident that these products would not have long to wait before 16e came along, and so would mainly be marketed as a stopgap or to traditional, fixed carriers in emerging markets with little need for the complexities of plug and play and portability. This has left them with 16d products that will not appeal to a broader base, especially if they are over simplified in areas like QoS, just as larger carriers are prepared to take a risk on 16d because of Rosedale 2; and with 16e products that are not sufficiently mature to succeed without the confidence stamp of certification.
The suppliers are adopting various strategies to reassure customers that there will be an easy route to 16e, once that is available - strategies that may not convince tier one operators with their focus on mobility and quadruple play, but that should be comforting to these suppliers' core target markets and to a broader base of tier two and three operators and WISPs.
Some are concentrating heavily on markets where the critical advance enabled by 16e wave 2 - nomadic, plug and play and embedded CPE - is less urgent. Hence a heavy focus by companies like Aperto and Alvarion on markets like India where the initial requirement is for fixed access and where truck roll is a minor issue - with Rosedale 2 as a promise that the systems will not be a dead end in future. Aperto claims to have transferred its whole PacketMAX architecture to 802.16e, allowing for a software upgrade since both systems are in TDD spectrum only, but other vendors cannot offer a smooth migration between their two products - especially if the first one is in FDD spectrum, since the FDD version of 802.16e is at least a year away.
Some are focusing on backhaul - both traditional point-to-point systems, in which companies like Redline excel; or selling 16d as an integrated network with Wi-Fi, to get round the low cost, portable CPE issue. Alvarion and some of the outdoor mesh specialists have been early into this approach, with Nortel and Motorola also looming.
Others, like SR Telecom, have taken the risk of ploughing significant development resources into a 16d platform that is positioned as more than a stopgap on the way to 16e, but rather as a fully functioning system in its own right, which operators will be able to migrate to 16e because of Rosedale and other advances, but when their mobility plans require that, not because it has run out of basic capability.
SR Telecom has taken this approach - ironically, given that in the early years of WiMAX it was one of the companies that aimed to go directly to 16e rather than incur a dual R&D burden, one of the most significant negative impacts, for smaller suppliers, of the decision to make 16d and 16e non-compatible. It changed its mind a year ago and received certification for its symmetry MX base stations and CPE last May (it also offers symmetry One, a pre-16e product). SR stakes its differentiation on implementing advanced techniques in both 16d and, in future, 16e, some of them - notably space time coding or STC, which improves indoor performance and plug and play capabilities - uniquely adopted by the company. Other important features are OFDMA modulation; sub-channelling to improve range, and hybrid ARQ (Automatic Retransmission Request) for better channel efficiency. It has also, like Aperto, put great efforts into QoS, supporting 1,000 classes of service including six channels for VoIP.
SR Telecom is highly aware that most large carriers - and its target base, including key customers Telefonica and Telstra, is more tier one than most specialists - require plug and play, self-installable CPE. It has taken the risk of ploughing big sums into developing this, even at a time of financial turbulence for the company, largely at the behest of key clients such as Telefonica for its wide scale Spanish roll-out. Now it hopes to carve out a leading position in indoor, plug and play CPE even before MIMO-enabled wave 2 16e makes this commonplace. STC is the main technique that SR claims puts it ahead of other indoor CPE makers - notably Airspan and Alvarion - and this is now supported by Intel in Rosedale 2.
Whatever transition strategy the WiMAX vendors are taking to weather the hiatus period before wave 2 certification - which itself will bring the onslaught of the tier one vendors - chipset support has been critical. For once, Intel has put itself in an early entry position with Rosedale 2, not just because of its dual-standard support but its early adoption of techniques like STC. Fujitsu and Wavesat have come out with chipsets for the often neglected 5.8GHz band (see Wireless Watch January 22 2007), and there is also increased interest in the 'ETG mode', which extends 802.16d into laptops and other applications and supports improved interoperability between 'd' and 'e' CPEs. Some chipmakers, notably Wavesat, have staked their short term business on this approach, arguing that 'd' is proven and will support most operators' business models for some years to come.
But the fact remains that indoor and, especially, mobile CPE at low price points will have to wait for key features, such as MIMO smart antenna arrays, that will only be certified in wave 2. And even then, for full wide area mobile systems, there needs to be considerable progress on creating a common specification for all aspects of mobility in the system, something even Sprint's three suppliers have not currently unified. Until that happens, a mass market will not materialize, and while there may be opportunities for specialized vendors and their customers, the big names will be holding their breaths. Motorola admitted that it does not expect to see revenues from WiMAX for at least another year, even though 802.16e is the heart of its infrastructure growth strategy, following its defocus on W-CDMA. Currently, WiMAX is an investment, not a profit, area for companies large and small, and the critical balance to strike is between investing enough to get a headstart when the market does take off, and being able to recoup that investment before the technology becomes commoditized. This is a balance that carries high risk for all participants, and may prove impossible to strike for some of the start-ups.
WiMAX and standards:
Speaking at the WCA conference recently, WiMAX Forum president Ron Resnick said that he had high hopes for WiMAX making progress in international standards bodies, a move that would lower the risk for smaller markets to adopt it. He said this month's International Telecommunications Union (ITU) meeting in Cameroon would see working party 8F discussing the approval of WiMAX as an IMT-2000 technology. This could make WiMAX officially part of the IMT-2000 family of standards at the Wireless Radio Conference in the second half of 2007. The significance, according to Resnick, is that "many of the smaller countries who really may not have all of the capability to do the due diligence to pick a technology, feel more comfortable if it is okayed by the ITU and it's implied it's a global technology".
Resnick also says that there has been progress with European standards body ETSI and with the 3GPP, positioning WiMAX to be a mainstream part of the pre-4G picture which will also include LTE. "3GPP approved WiMAX at its last group meeting. It approved the two technologies and wrote an inner specification for it. It's all done," he said in an interview.
Rosedale 2
Rosedale 2 was launched in mid-2006 and is now being incorporated into products by many WiMAX supporters, stealing an unusual march, for Intel, on more specialized chipmakers in WiMAX.
The product is bit compatible with its 16d-only predecessor, Rosedale, and is the first to integrate Intel’s global WiMAX radio chip. It will be used initially in residential modems, particularly to boost the slow progress towards certified plug and play devices - 95%-plus of WiMAX CPE remains outdoor or operator installed. It could also find its way into pico base stations, mirroring the 3G community's new interest in tiny indoor or outdoor access points to support indoor penetration and mesh-type configurations - a trend led in WiMAX by UK chipmaker PicoChip.
Intel's next product to appear in devices will be the single-chip radio called Ofer-R, which supports both Wi-Fi and WiMAX and is geared initially to laptops, and then to other portable devices such as MP3 players.
rethinkresearch.biz
Cheers! TM |
