The image of WiMAX as being a ‘grown-up Wi’Fi’, as it was often called in its earlier days, was once a definite plus. It made a hitherto obscure technology readily understandable and generated excitement on the back of the boom in all things WLan. Now this association is becoming something of a curse. The focus of the WiMAX community is shifting rapidly towards the licensed spectrum profiles and mobility, and so to a platform that must be truly carrier class, and necessarily far more complex than Wi-Fi. But there is still a belief that this platform can be created, and certified for full interoperability, in a single relatively simple process that would actually only be appropriate to low cost, unlicensed band products. The impossibility of delivering a sophisticated mobile broadband platform, with a development and certification process as simple as that for consumer access points, has created a gulf of understanding that is now being used in anger by WiMAX’ detractors, and is causing genuine confusion in the market.
There is a real dilemma for WiMAX suppliers. Should the process, already heavily criticized for slipped deadlines, be delayed further to make the ensuing products carrier class? Or should equipment be rushed out to stave off the critics, but with only basic capabilities certified?
The answer is both ‘ there is a role for basic and for complex gear in different sectors, but it must be very clear to the market which is which, and that the latter will take longer to be fully standardized. This in turn requires a more complex understanding of the testing and certification processes behind this multi-layered agenda.
Some of the problem lies in the way that the WiMAX agenda has changed over the years, so that now it has a hugely broad remit, from unlicensed WISP equipment similar to Wi-Fi, to full mobile carrier networks comparable with 3G. Wi-Fi is a (relatively) simple technology in terms of functions, 3G is not, but it is only in the WiMAX world that the two types of network can be confused. From 802.16’s early roots in the unglamorous backhaul market and the unloved wireless local loop, the attention of Intel and the PC agenda shifted the focus sharply towards non-line of sight, portability and, indeed, a metro area version of Wi-Fi, and then on to mobility. However, there was still scope for a straightforward and logical standards progress. A fixed/portable standard would be defined and certified and then a subsequent effort would lead to a mobile version, both sharing the same physical layer for smooth transition.
The process of creating the platforms would be linear ‘ like Wi-Fi’s, where the original 802.11b provided a fully formed technology, not an interim technology waiting for the real thing to come along, and subsequent enhancements such as higher speed versions were added in later waves of development and testing that still retained (with the exception of the 5GHz 802.11a) the same fundamental technology and were targeted at roughly the same markets ‘ consumers, WISPs and enterprises.
The market has widely expected the same experience in WiMAX ‘ waves of development that would enhance but not fundamentally change the original technology, and a first certification wave that would produce, in short order, a fully fledged solution ready for immediate uptake.
The reality has become entirely different, largely because of two factors. First, the mobile technology is now entirely different from the existing fixed 802.16-2004 platform, having adopted the Samsung-inspired SOFDMA physical layer. Therefore development and testing of 802.16e equipment is not a linear extension of the work on 802.16-2004. Second, the critical target market has shifted towards tier one and two carriers looking to add or enhance mobile broadband services. This has attracted in major vendors, such as Motorola and Lucent, and caught the interest of major potential customers, creating a far larger possible market that was dreamed of even two years ago. But this has also vastly increased the complexity and quality of products that will be required, and so put new pressure on the R&D and testing processes. The 3G technologies were not created or certified in a matter of a few months, and fully fledged mobile broadband WiMAX will not be either. But because of the early expectations and positioning, this is causing disappointment and confusion.
The vendors are well aware that the goalposts have changed and that the real money will be made in the carrier class market, with the more traditional WISP sector, with its relatively simple requirements and unlicensed or low cost spectrum, somewhat sidelined. Recent milestones such as the first WiMAX trials by tier one carriers, or the incorporation of 802.16 by Alcatel and Lucent (both Alvarion OEM partners) into their IP Multimedia Subsystems, reflect the effort and dollars that are going into making WiMAX carrier class in the same sense as 3G. But the market ‘ investors, commentators and some service providers ‘ still clings to the perception that the critical process of certifying WiMAX equipment will be relatively simple, and that as with Wi-Fi, there will be a single linear process for each frequency profile, leading to a deluge of uniform and fully formed equipment hitting the market in shiny, carrier-ready form around the turn of the year (not that the Wi-Fi process was ever that simple either, but at least, once the politics of agreeing standards were over, the market requirements were fairly straightforward).
WiMAX vendors are put in a dilemma by this situation. The WiMAX Forum is still committed to delivering on its schedule of seeing the first certified standard equipment in the market around the end of 2005. But the speed of the process requires that these first tier products will only be tested and guaranteed for the bare essential, the air interface. The standardized functionality will be low level and upper layers will remain absent or proprietary, limiting their value. This will also limit, we would assume, uptake of these products, keeping volumes small and delaying the hoped-for price reductions that will spring from a volume market (bad news, mainly, for small operators that have invested in pre-WiMAX gear and are counting on such price reductions to sweeten their business models early in 2006). Therefore the main reason to publish certification for this first stage is for publicity value and a statement of credibility about commitment to standards. Most existing suppliers will find that their existing semi-proprietary gear is more sophisticated and functional than their certified offerings, and so it will be hard to find a real market for it.
Yet if they delay until the certification moves on to more advanced iterations, they will be accused by rivals of falling behind on their schedules.
Alvarion, one of the suppliers caught in this dilemma, believes the misconceptions would be largely swept away with a different view of the testing and certification processes. It has come up with a pyramid image (see diagram), which will be adopted by the Forum, to replace the traditional view of WiMAX certification as a linear series of waves. Instead, the move towards full mobile broadband will be about a basic, broadly supported but fairly low functionality foundation, on which more and more specialized and complex layers will be built, reaching the apex of full mobility (and future enhancements in future as we move towards ‘4G’). Rather than releasing commercial and certified equipment for each layer, some vendors will choose to do two or more layers at once, depending on their market focus. Motorola, for instance, plans not to join the certification process until the fourth, apex layer, because it is only interested in the mobile market. It will, of course, have its offerings tested for compliance with the basic air interface, but it would be a waste of its money and efforts to release products without all four layers, since they would be inadequate for its chosen target base.
Alvarion is in the same position, although, since it does have strong presence in the fixed wireless market, it plans to leap on to the pyramid at an earlier stage, in the second layer, where WiMAX gains certification for key functions in the outdoor CPE market, notably quality of service. In wave one products, QoS will be included in a pre-certified form only.
‘Alvarion will be part of wave one but we won’t seek certification,’ explained global head of marketing Carlton O’Neal. He gives two reasons. One is the need for relatively small players to prioritize their resources and Alvarion feels its time and money is better spent on finishing the current release of software ‘ release 2.5 ‘ for its BreezeMAX 3500 pre-certified WiMAX range and shipping that to customers, than on attaining a certificate that merely proves BreezeMAX complies with the air interface.
The second is that, O’Neal believes, releasing an officially certified product would create false expectations among existing BreezeMAX customers. Since the market is still largely wedded to the linear progression image of WiMAX evolution, there is a widespread assumption that the first WiMAX Forum Certified products will be a step ahead, in functionality terms, of current pre-certified offerings. Many service providers, for instance, are hoping that a certified product from Alvarion or Airspan, the two vendors that have so far shown off indoor customer premises equipment (CPE) units, will include those indoor facilities. In fact, as the pyramid indicates, first wave certification will apply only to the air interface.
‘We have 130 customers for BreezeMAX,’ said O’Neal. ‘We are working on release 2.5 of the software, which includes the capability to enable the self-installed indoor CPE. That is what the customers want first and we won’t delay that to prove we can work over the air. Also, wave one will not have the functionality level we are currently offering in BreezeMAX. Customers will want certification but they will expect that to equate to release 3.0, which is coming next year.’
In other words, the standards will not catch up with real world, pre-certified equipment in terms of scope until the first half of 2006, with wave two, the first to test real commercial functionality, and wave three, which extends that to the vital systems with indoor, portable CPE.
So what is the point of wave one at all?
There is a clear point to it from the point of view of the WiMAX Forum. To merge waves one and two might make the first certified equipment more attractive to serious operators, but it would delay availability by several more months at least. This would be a major credibility and public relations blow, given that the Forum has already been forced to lengthen its timescales for certified equipment availability already. Original estimates of these timescales, made back in 2003, were wildly optimistic, and the WiMAX delays have been no more serious than those that generally afflict new wireless technologies, but they have been frustrating to vendors and some early customers and have proved valuable fuel for WiMAX detractors. Delays also raise some serious issues, particularly for the mobile WiMAX standard. For parties whose interest lies in positioning 802.16e as an alternative to HSDPA and CDMA EV-DO Rev A, it is vital that the market believes the newer technology will not also be lagging a year behind the other mobile broadband options.
The other group that can benefit from supporting a fairly basic first set of specifications is the new entrants to the market. Companies that are moving from a different sector, such as backhaul or wireline access; that do not have a significant base to support in BWA; and particularly, those targeting WISPs with fairly simple equipment requirements, perhaps extending an outdoor Wi-Fi system, will at least be able to get a toe in the market and make some early, if low level sales (though this argument would have been strengthened had the 5.8GHz profile, most used by smaller WISPs, been addressed first in wave one, rather than licensed 3.5GHz).
For companies that have sophisticated broadband wireless systems already in the market, there is a strong motivation not to break the revenue momentum behind those products while they are still selling, and not to excite customers over a standard that will not yet meet their needs. Indeed, provided the user base is still willing to purchase pre-certified (or, in the case of some suppliers, proprietary) gear, based on the promise of a migration to full standards, there is an active interest in not leaping too early into the price war that standardization inevitably brings. As long as high level functions are not part of the standard, they can act as a differentiator, and a margin enhancer, for those vendors that offer them on top of a standard base.
In our view, wave one is vital to the chipmakers, but to everyone else it is important as a credibility boost and a stepping stone rather than because it will bring real WiMAX systems flooding on to the market. As O’Neal describes it, it is like the foundation of a house ‘ essential, but you can’t live in it. Some companies will therefore emulate Alvarion and not seek certification for this level of testing ‘ others will feel they need the certification to reassure customers, but will still have to add significant additional functionality outside the basic platform to make real sales. At which level a supplier jumps on board will depend on their market profile ‘ how far they feel they can attract the confidence of customers even without early certificates ‘ and on their target market. Vendors keen to focus on the 2.5GHz licensed spectrum profile, newly exciting because of the major planned build-outs by Sprint Nextel and Clearwire and Bell Canada/Rogers, will see little point in taking the effort to certify even at wave two, since all these deployments will depend on indoor and portable CPE. Alvarion itself will seek certification for its 2.5GHz products at the wave three stage, which could mean a wait of nine months or more.
WiMAX has found itself in a difficult position. Its scope and ambition have vastly expanded in the past two years. From being a more economically attractive version of traditional BWA niche products, or a technology to extend the promise of Wi-Fi in unlicensed bands, it has become a truly mobile broadband platform with appeal to large carriers and the chance to support new markets and players. But it has been stuck with expectations of a certified product roll-out schedule that is wholly unrealistic for such a complex requirement. The WiMAX Forum needs to ensure, with tactics such as adopting the Alvarion pyramid image for its testing process, that expectations are realistic. That may hand some short term weapons to enemies by disappointing the most optimistic hopes of when WiMAX gear will be available in volume. But the counter-message must be that, when wave two, three and four equipment does appear, it will be carrier class, as a rushed job could never be. Being first to turn up at the labs in Cetecom, or first to put a WiMAX Forum stamp on a box, is good publicity but little more. The operators are looking for well rounded systems with functionality that is better, or at least equal to, proprietary choices. They are likely to be prepared to wait another six months for that.