Standard bearers

After a year of vicious disputes in the wireless arena, rival bodies, technologies and chipmakers are finally edging towards achieving consensus in the sector

  • E-Mail
By  Peter Branton Published  June 4, 2006

|~|analyst66body.jpg|~|Wireless standards, particularly the WiFi v WiMax debate, have been the subject of much discussion in the industry.|~|Last year was characterised by vicious disputes over wireless standards, the worst of which, for UltraWideBand-based 802.15.3a, led to the project being abandoned altogether. This has raised questions over whether bodies like the IEEE (Institute of Electrical and Electronics Engineers) are still relevant, with their processes and political in-fighting lagging well behind real market demand, and encouraging vendors to leap ahead of industry specifications. Such questions, and the threat to IEEE platforms from a rejuvenated cellular agenda and other quarters, seem to be encouraging vendors to take a more cooperative line. The 802.11n fast WiFi standard-to-be, once also divided between two warring factions, has now achieved a combined proposal, which could lead to a final specification by the end of this year. And now the two groups fighting to provide the core technology for 802.11s, the future standard for WiFi mesh, have also merged to create a consensus proposal, which has been adopted by the IEEE as a draft. If these mesh groups, led by Nortel and Intel, had remained in conflict, it could have delayed the standard to such an extent that vendors and operators in the attractive metrozone market would have lost confidence and turned to other networks such as WiMax or even cellular. With this in mind, the mesh specialists have reached a deal, and are aiming to create a broad-based standard, on which more specific profiles will be built — in the same way as WiMax. This will enable 802.11s to address various markets, not just its initial enterprise campus focus but also the metrozone and other future applications. It will also support the addition of other radio technologies into the platform, including WiMax. For all the apparent harmony though, there may be more disputes down the road, a trend already threatening to delay the apparently harmonised 802.11n further, encouraging vendors to launch in pre-standard mode. Although the 802.11s standard for WiFi mesh will not be ratified for at least another 18 months, the IEEE task group charged with creating the specification will at least be working from a single starting point now, avoiding the factional fighting that plagued the 802.11n fast WiFi efforts until recently, and which killed the 802.15.3a UltraWideBand initiative altogether. Such delays and uncertainties would have been highly damaging in the metrozone space, the great hope for all the mesh suppliers right now, since they could have divided the market in two, with budget-constrained cities facing the risk of choosing equipment that might be non-standard and therefore have a short shelf life and no future interoperability with other equipment. The 802.11s platform will standardize the currently proprie- tary protocols that route traffic around the mesh and will support mobility and broadcast/ multicast/unicast signals. The WiFi Alliance, or a separate body, is likely to take respo- nsibility for creating different profiles or subsets of the full standard, as has happened in WiMax, and this activity could also open the way to including other networks in the mesh alongside WiFi. There is a high degree of interest in combined WiFi/ WiMax metrozones, with the latter providing backhaul and premium services, perhaps in licensed spectrum. Profiles are likely for different types of mesh behaviour and this will be important to make the standard meaningful. There has been much debate over how all-encompassing 802.11s should be — for instance, whether it should cover both indoor and outdoor usage. Profiles allow a potentially unwieldy standard to be subdivided into more manageable chunks that are optimised for certain applications and can relatively easily unite the vendors in that sector and so create volume economics. Examples could include profiles for single and switched multiple radios; contention-free sectorised mesh; and various frequencies such as 2.4GHz, 4.9GHz, and 5GHz. The original 15 proposals submitted to the 802.11s task group were quickly whittled down to two, under the auspices of two coalitions. One was the Wi-Mesh Alliance (WiMA), created last year and led by Nortel, with supporters including Accton, ComNets, InterDigital, NextHop, Philips, Extreme Networks, Mitre, the US Naval Research Laboratory, Swisscom Innovations and Thomson. The other was the SEEMesh group, backed by Intel, Nokia, Motorola, NTT DoCoMo and Texas Instruments. The urgent need to maintain confidence in mesh clearly outweighed the different agendas of the two parties and drove a fairly quick compromise. This has been portrayed as a response to the ‘threat’ that WiMax would pose to metrozone mesh — for instance, Motorola, which plays in both technologies, has produced models that show WiMax as a more cost effective way to blanket a densely- populated city than WiFi mesh. And WiMax has clear advantages such as the choice of licensed or unlicensed spectrum for different required service levels. However, we believe WiMax and WiFi will more commonly combine than compete in this particular market, especially with the rapid appearance of dual-mode devices for the two radios, as previewed by Intel last week. The real danger to mesh sales in the city environment is the fight back from cellular operators; offering better priced more powerful 3G and 3G+ services that, they argue, make the mesh a waste of city money. Not that the world of WiFi is entirely harmonious. True, the two factions that were fighting last year to provide the 802.11n fast WLAN standard did come to a consensus — albeit one largely forced upon them by the major WiFi chipmakers — Intel, Broadcom, Atheros and Marvell — effectively threatening to support a breakaway movement and bypass the IEEE were their ideas not adopted by 802.11n. This was a real threat — in the 802.15.3a stalemate, Intel and Texas Instruments did remove their WiMedia Alliance from the IEEE altogether and gained standardisation from another body, ECMA, effectively killing the IEEE effort. In 802.11n, a consensus proposal was adopted, however grudgingly in some quarters, but that has not stopped the bickering over certain aspects of the platform, notably its support for optional wide (40MHz) channels. This echoes previous tussles over apparently obscure but politically charged technological topics — such as last year’s epic about what really constitutes MIMO (Multiple Input Multiple Output, the smart antenna approach at the heart of 802.11n). As with that debate, the latest one is as much about politics as wireless technology, and pitches the two claimants to WiFi MIMO pre-eminence, Airgo and Atheros, against one another. Airgo was the big loser from the acceptance of the Intel-inspired 802.11n proposal, which was supported by Atheros. Airgo was first to market with MIMO chipsets in the WiFi arena, and has signed up OEMs such as Linksys and Belkin for its ‘True MIMO’ devices. These claim to be ‘pre-802.11n’ but are in fact proprietary performance boosters for current WiFi standards, since 802.11n has not yet been ratified or fully defined. However, they do bring ‘n’-class performance — well over 100Mbps — to current WLAN ranges, provided Airgo chips are used at both access point and client end. Atheros has a similar offering, though one dismissed by start-up Airgo as ‘not really MIMO’. Whatever the rights and wrongs of that intellectual debate, Airgo is now faced with the prospect of having to overhaul its technology to support 802.11n, rather than its chips becoming the heart of the new standard, as it had previously hoped. For a start-up, this will entail unwelcome costs, plus the loss of the market head start it had created. Little surprise, then, that there are some sour grapes, with Airgo claiming it will not build a chipset for draft one of 802.11n because it will not be good enough. It also claims the draft has problems that will delay the emergence of the full specification and subsequent products —a point of view denied by Atheros, which said the newly released draft 1.0 has hardly changed from the original draft, demonstrating its innate stability and closeness to the final stages. Airgo’s complaint is that the current draft would degrade or even disable existing 802.11b/g networks in the 2.4GHz band, and it implies that its rivals are deliberately seeking to push a standard that does not co-exist with current products in order to steal Airgo’s market share in the fast WiFi sector. This is overwrought — Airgo’s share of the WiFi chip market, despite its early moves in MIMO, is dwarfed by those of Atheros, Broadcom and Intel. But the first two, plus Marvell, have shown off products based on the 802.11n draft and aim to release chips long before the final standard is ratified, so this really has become a game of timing — one where Airgo may have overplayed its hand, and where Broadcom, if past history is anything to go by, will get it just right. The authors are Caroline Gabriel, Experture Expercycle analyst, and Kenn Walters, Experton Group partner advisor and Tecom research director||**||

Add a Comment

Your display name This field is mandatory

Your e-mail address This field is mandatory (Your e-mail address won't be published)

Security code