Rethink Research
- Wi-Lan likely first of many specialists to leave WiMAX equipment market
- Will now focus on patents business but may need to build on Cisco deal
- Other companies will find it hard to survive without large partners
- Patents remain issue says Forum, but 802.16e testing the most pressing issue
- Regional expansion sought for testing facilities
In a standards-based equipment market, volume is key to competitiveness, and smaller suppliers will only survive if they have significant intellectual property over and above the standard. The Wi-Fi market is in the process of consolidating and it will soon be WiMAX' turn. We have already seen Proxim and Terabeam merging, and now Wi-Lan is staking its future entirely on its patents portfolio, and seeking an acquirer for its equipment business. With the certification process for 802.16-2004 fixed WiMAX kit speeding up, and mobile 802.16e on the horizon, this will certainly be the first of many upheavals in the once cosy world of broadband wireless vendors.
Wi-Lan:
Canadian Wi-Lan has had one of the most tumultuous histories since the broadband wireless market started to converge on Wi- MAX. It has been in financial difficulties, and undergone a series of reorganizations, over the past two years, but has clung to its claims of significant intellectual property holdings in Wi-MAX and other OFDM technologies such as 5GHz Wi-Fi. Wi-Lan believes its portfolio of patents, including its core WOFDM patents and 17 patents and patent applications acquired from now defunct WiMAX pioneer Ensemble Communications in May 2004, are necessary for the implementation of devices using the IEEE 802.16 and ETSI HiperMANas well as Wi-Fi variants 802.11a and 802.11g and ETSI HiperLAN/2.
These claims led it to take the bold move of suing Cisco, and it gained licensing agreements with its key technology partner, Fujitsu Microelectronics, with whom it codeveloped the Fujitsu WiMAX system on a chip.
However, despite some licensing wins, Wi-Lan did not succeed in gaining the Qualcomm-style universal recognition of its patents that it had hoped for. Indeed, it was widely rumored that Qualcomm was interested in acquiring Wi-Lan as part of its bid to assemble its own OFDM patent arsenal, but instead it opted for Flarion, a move that suggested to many that Wi-Lan's IP was less attractive than the company had hoped.
Cisco did show some interest, and may still be Wi-Lan's main salvation, as the networking giant seeks to flesh out its mobile IP and OFDM patents store in an intellectual property stand-of that looks increasingly likely with Qualcomm. In December Cisco came to a settlement with Wi-Lan, and licensed all the smaller company's patents and, significantly, acquired some of them for itself, in the WiMAX and smart antenna areas. Like Qualcomm with the acquisition of Flarion, Cisco is building up its patent stores in key technology areas, and more of Wi-Lan's intellectual property may also be snapped up.
Specialists survive?
Now the company plans to phase out its struggling equipment making arm and focus entirely on its patents - a move that we can expect to see from other smaller vendors. Unless they can gain significant OEM partnerships - as Alvarion has with Alcatel and others, and Airspan with Ericsson and Nortel - these specialists may find it hard to compete with the giants. As it is, most of the current OEM deals are in the fixed WiMAX world, a relatively niche market for backhaul and traditional BWA, that is of limited interest to the giant vendors and so could continue to support growth for the niche players as the better economics of WiMAX kick in. Redline, for instance, has a strong track record in backhaul on which it will continue to build in the fixed WiMAX market.
In the 802.16e market, focused on consumer access, triple play and, eventually, mobility, the large suppliers that are most serious about WiMAX, such as Nortel, are set to develop their own products, loosening their ties with the specialists. This will create a degree of shake-out in the 2007-8 timeframe, but at least those companies with a strong partnership for 802.16-2004 have a breathing space to extend their business in other directions - especially though licensing, as we have seen SR Telecom doing - or secure an attractive acquisition. For Wi-Lan there was no such luxury, as its cashflow was weak, and it had failed to gain a big name licensee. It will now phase out the equipment business by the end of April, saying this arm was not profitable and the company has been "unable to sustain the costs of development expense of its products while also growing the intellectual property business".
In August, it had already split off the IP unit as part of a move to take a more focused approach. This had earlier led it to place most of its Mobilis high speed transport- oriented technology into the hands of a Korean partner.
Wi-LAN's Mobilis pre 802.16e technology
It said it is now looking for potential buyers for the equipment business and is "working with several other companies to form new approaches to market its existing IP systems, patent portfolio and expertise, and to develop new IP products for the wireless fixed and mobile markets". It is to be hoped that the new strategy does not merely turn Wi-Lan into a patent hoarder seeking to take aggressive action against product developers, an increasing common situation in the wireless world and one that merely serves to confuse the market and delay systems roll-out.
Intellectual property issues:
The intellectual property issue remains one of the uncertainties surrounding the progress of WiMAX. As well as the lurking threat that Qualcomm may claim patent rights to some aspects of 802.16e systems, particularly in the mobile IP area, the policies of the IEEE and WiMAX Forum on licensing have always been vague and unenforceable, especially with the accelerating trend for standards bodies to take existing de facto standards on board in order to speed time to market, rather than starting from scratch. Many players - including a group of large cellcos led by Vodafone, which recently achieved a significant shift in EU rules on patent claims - are calling for all intellectual property claims to be declared up-front during standards development, or relinquished, to avoid the rising risk of "patent ambushes" when claims emerge only when a standard has been approved and is gaining market momentum.
Jeff Orr, executive at the WiMAX Forum and head of marketing, agreed the process is complex: "We need 'reasonable and non-discriminatory (RAND) policies on intellectual property but if a standards body has the intention to use a de facto technology, RAND is hard to implement". This could clearly apply to Qualcomm and its likely attempt to push its newly acquired Flarion Flash-OFDM technology as the basis for the 802.20 standard for mobile broadband, a potential alternative to 802.16e.
Progress on certification:
Patents are not the only complex issue facing the Forum, of course, as it races to meet the market's expectations for certification of 802.16e equipment. Despite the long and often badly communicated process of getting the first 802.16-2004 gear out of the certification labs, run by Cetecom in Spain, the process does now seem to have gained momentum for the fixed wireless products, and confidence is growing. It is very important that this is maintained for the more critical 802.16e standard, and the Forum is showing signs of rising to that challenge.
Part of the problem for the initial testing effort was the expansion of WiMAX' remit and market profile during the run-up to its standardization. From being a somewhat obscure technology for traditional fixed wireless and backhaul, it had taken on the mantle of being a mainstream consumer access and mobile network capable of delivering a triple play service for a major operator, and so had gained huge new political significance and the interest of many large players. This meant that the scope of the WiMAX Forum's work had to stretch almost beyond its resources For instance, Orr says far more products joined the queue for first stage certification than had been expected, and Cetecom has had to adjust its resources to meet the demand and ensure that companies can get into line within a few weeks.
So far, a handful of 802.16-2004 products for the 3.5GHz FDD (frequency division duplex) profile have been certified of a total of over 30 in line, and the first 3.5GHz TDD systems will soon emerge too. The certification of 5.8Ghz fixed WiMAX gear will take place in parallel and will start as soon as a testbed of at least three products is checked into the lab to form the basis for the testing. Defining the 5.8GHz process should start within a few months and the actual certification will be quicker than for 3.5GHz, as areas like protocol conformance are common across different bands.
The Forum is now looking for ways to expand its laboratory facilities, either by using other Cetecom locations and/or signing new partners. This will be essential to cope with the addition of 802.16e to the process - validation of test processes for the initial 802.16e profiles, which are currently under review, is expected to start around August, with the first certified products emerging at the turn of the year, though some vendors like Motorola will launch pre-certified gear that is built to the 'e' specifications but only tested unofficially.
Addressing new geographies:
Additional lab capacity will also be needed as WiMAX continues to expand its remit - new profiles are being considered, particularly to meet the frequency allocations in special case markets like Japan and Russia. The aim, according to Orr, is to have mainstream testing centers for 802.16e in the three key geographies - Europe, Americas and Asia-Pacific - and some localized labs for countries with specialized profile require ments, notably Japan, China and Korea. The Forum is facing the classic dilemma of a would-be global wireless network reducing the standard to just a few profiles encourages uniformity, volume and attractive pricing, as well as appealing to operators with support for global roaming; but some important markets will not harmonize their spectrum policies for years to come and must not be excluded. This dilemma afflicted 5GHz Wi-Fi but was relatively easily solved since there was only one major profile affected and many countries are moving into line with ITU guidelines on license-exempt 5GHz rules. For Wi-MAX, with profiles in licensed bands that are different allocated and governed in different regions, the challenge is far more complex.
To try to keep important economies, despite their peculiar spectrum maps, in the WiMAX fold, the Forum is recruiting 'champions' in key growth markets to promote the technology with operators and regulators. Initially these will be in China, India, Brazil and Russia.
The WiMAX certification process and cycles - Source: WiMAX Forum
