The Uncharted Road to Mobile WiMAX
>> Is Fixed WiMAX DOA?
Wireless Watch
August 30, 2005
tinyurl.com
Alvarion attracted considerable publicity for failing to be among the first vendors to deliver equipment to the WiMAX Forum labs for testing. However, arriving early has more marketing than real value, and the first wave of certification and interoperability testing is unlikely to start until October.
This makes it difficult to fulfill the promise of delivering kit to market by the end of this year, but even more importantly, the long process is increasing vendor interest in moving straight to the mobile 802.16e standard. This was initially seen as an option for companies whose customers are only focused on mobile services, but now some suppliers are seeking to go straight to the 'e' standard even for fixed offerings.
SR Telecom is one example of this, while others such as TeleCIS remain loyal to the idea of providing an enhanced version of the current platform and adding feature such as sub-channelization, to give the initial technology a reasonable lifespan before the difficult upgrade to 802.16e becomes necessary. Increasingly, the question is not how quickly vendors can achieve 802.16-2005 certification, but whether they should bother at all, or save their resources for 802.16e, to ensure a harmonized standard whose own testing processes are not backed up behind those of the fixed platform.
But certification testing is not a PR matter. It is complex, technical and full of unexpected surprises, and the focus on Alvarion's decision not to turn up at the lab until September or October has obscured the far more important issues that surround the start of 802.16-2004 certification – how efficient will the process be, and how long before the market has a good choice of certified, well priced equipment; how important is fixed wireless testing anyway, with most of the big names waiting for mobility; can the equipment that is certified deliver on its promises?
Timing
All these are real issues that will affect the progress and success of the WiMAX technologies. The timing of vendors' arrival at Cetecom is not highly significant, provided all those planning to deliver fixed equipment are there by year end.
The timing of getting certified products to market will be far more indicative of any problems and risks for WiMAX in its early adoption stages. To paraphrase Oscar Wilde, it's not your first test that's important, but your last – the one that gets you certified and out on the market. If Alvarion does have problems with its kit that have forced it to delay delivery, as its rivals would have us believe of course, these are unlikely to have any serious impact on WiMAX adoption.
The first round of testing is focused only on small players (with a couple of exceptions, notably Siemens), and among these, Alvarion is the giant, with greater engineering resources than its competitors to speed its way through the certification process, making any tweaks that are required to achieve full compliance. The Israeli company's own explanation of why it will not show up in Malaga until the fall is that the full testing process will not get going until September anyway, with Cetecom currently working on getting the testbed finalized, a task that will not be finished until October, and so there is little point in arriving early.
Achieving a Testbed
The real reason why the Forum might be keen to get all the vendors there at the same point is to ensure that there is a quorum of participants in each interoperability test – in other words, with 3.5GHz the first profile to be tested, there need to be at least three suppliers present with 3.5GHz kit for TDD (time division duplexing) and three for the FDD (frequency division duplexing) variant.
With most of the tier one telco equipment vendors sitting out the first round, there is a risk that 802.16-2004 testing will be based on a very small pool of equipment, which could dent confidence of potential buyers and make it even more likely that they will wait for the 802.16e mobile release before making major purchases.
This, in turn, would make it hard for the mainly small suppliers of fixed WiMAX kit to achieve the volumes and economies of scale necessary to deliver on their price promises. Most equipment majors remain focused on 802.16e, and are bringing increasing pressure to bear to ensure that certification of that standard – which should be ratified this fall – will not be held back by any delays in 802.16-2004. Motorola will not seek certification for its planned Wi4 range until 802.16e is in place; Nokia is focused only on mobility; and other majors such as Alcatel and Ericsson will rely on third party equipment until the mobile stage, when they will develop their own gear and step up their efforts a notch.
SR Telecom and 'e-'
It is not just the giants that are hanging on for mobility. Some companies that are entirely focused on portable or nomadic systems, such as Navini, will wait for 802.16e, as will others such as SR Telecom. SR believes 802.16e will be more appropriate to the needs of larger roll-outs of several hundred base stations and more – a scenario where WiMAX remains virtually untested as yet, with most pre-standard deployments still limited to a few hundred subscribers.
SR, though it has struggled with financial difficulties over the past year, can at least claim to be a big carrier supplier, with three networks of over 100 base stations in place using its current technology, which has a migration route to 802.16e. SR's technical pedigree comes from acquiring the assets of AT&T's broadband wireless Project Angel, and its largest customer is Spanish telco Telefónica. Another large user is Globalcom in Lebanon, which has 20,000 subscribers on its system.
Few existing BWA suppliers apart from Alvarion and Motorola can boast really large systems in commercial use. The decision by SR is interesting, and echoes that of some far larger carrier-class vendors – the company is not focused only on mobile systems, but is taking the line that 802.16e will be a better option anyway, even for fixed and portable networks. SR believes that vendors that comply just with the initial specifications for 802.16-2004, and no more, cannot deliver a carrier class solution, since functions such as space-time coding and subchannelling are not mandated.
In other words, being first to certification does not make a vendor competitive for large customers. This is an argument that is gaining increasing weight, since adopting 'e' from day one will save the vendor, and its customers, making a major upgrade in order to support mobility one day, and 'e' brings various efficiencies to fixed systems too, from its distinct physical layer, based on the Scalable OFDMA implementation pioneered in the South Korean Wi-Bro technology. Also, there is an increasing sense that most of the industry's resources will go into 802.16e, as well as most of the major deals.
e-minus vs d+
By promoting a so-called 'e-minus' approach – aiming to deliver the benefits of the newer standard at an early stage, even before full mobility is supported or demanded – SR and others are seeking to draw in even the fixed WISPs that might have taken little interest in the mobile platform.
This contrasts with the 'd-plus' approach of companies like chipmaker Wavesat – adding as much functionality as possible to the current technology in order to support portability and other features, and prolong the life of the fixed WiMAX equipment. Both points of view have their technical pros and cons, and 'dplus' at least has the virtue of standards that are already ratified and a headstart on the certification race. But the weight of big vendor support and R&D is being put behind 802.16e – weight that may grow heavier following Qualcomm's acquisition of Flarion, a move that could galvanize anti-Qualcomm factions into action.
This suggests that the e+ route will be the safer one in terms of investment protection for vendors and users, even though the precise technology remains untried. That fact will focus the industry's attention firmly on South Korea next year, as the major Korean operators roll out systems based on Wi-Bro, the technology at the base of 802.16 and therefore the best pre-standard indicator of how it will perform. The vendors are already testing their Wi-Bro gear and this may bring valuable intelligence to 802.16e testing that could streamline it rather more than the 802.16-2004 process has been so far.
It will also, as Alvarion's marketing head Carlton O'Neal points out, put Samsung, the main Wi-Bro vendor, in the driving seat in terms of mobile WiMAX certification. So, arguments about which vendor got to Cetecom first are totally overshadowed by the bigger question – is there much point testing for 802.16-2004 certification at all, or would suppliers be better to save their money and efforts for the next generation? As we have seen before, there will be a market for fixed WiMAX, in the traditional broadband wireless base of 5GHz WISPs and rural telcos, as well as for backhaul.
Ironically, this market could be boosted in terms of volume by the cut-throat price competition that is likely to develop, as vendors fight for a smaller base than was previously envisaged for the first 802.16 platform, seeking to gain return on their R&D investment. This should stimulate WISP uptake at an early stage, but this looks likely to remain a market primarily for traditional suppliers. The really rich pickings will come with 802.16e, and that will not be purely a mobile standard any more, but should take on the mantle of 'the real WiMAX'.
Addendum: The Testing Waves
The first stage is to validate the test itself, a process in which many vendors will take part and which will take until October. After this, there will be four 'waves' of testing:
1.) Time division duplex kit in 3.5GHz band.
2.) FDD kit in 3.5GHz.
3.) 2.5GHz equipment.
4.) 5.8GHz equipment.
The last two are expected to be completed around May 2006. <<
- Eric - |
