Will pressure to speed up 802.11n wreck standards process?
Pre-empting the pre-empters
This week sees a major IEEE meeting in Vancouver, Canada, with high hopes that the standards body will accelerate progress on several of its key specifications. There is particular pressure to speed up the development of the 802.11n extension to Wi-Fi, which will support speeds of up to 108Mbps, since vendors are already pre-empting this with proprietary fast WLAN products. However, this is raising fears that the agenda for such standards is falling entirely into the hands of the suppliers and the vendor-driven Wi-Fi Alliance, which may be forcing the pace at the expense of the quality and independence of the IEEE process.
The IEEE’s 802.11 Wi-Fi group is expected to issue a request for proposals for the 802.11n extension to the standard this week. This will mark the official authorization of the project and lead to the usual dogfight over which of the technologies complying with the initial specification will be incorporated in the standard.
This would indicated a period before final ratification of about two years, which would be significantly shorter than previous Wi-Fi variants – although there have already been discussions of high throughput amendments stretching back for another two years. Market demand for fast Wi-Fi and moves by vendors to turbocharge 802.11g and even devise pre-standard 802.11n are all making even this timescale seem overlong.
However, the 802.15.3a experience has pointed to the dangers of streamlining the standards process too much. And there are very weighty issues to be considered before ‘n’ is finalized.Its key difference from existing Wi-Fi variants is that it measures performance where the media access control (MAC) layer meets the higher layers, not at the PHY (physical layer).
This should ensure a far smaller gap between theoretical and real world data rates. But it also raises the likelihood that both the MAC and PHY will need to be tweaked significantly, and also raises the possibility that whole new ones could be considered, making the task a complex one.
Taskgroups to date have sought to limit the scope of the work, partly in order to speed the process and also so as not to lose valuable work done in the past. There is a baseline specification to which 802.11n must conform, which includes all existing 802.11 functions. Nothing can be removed and only mechanisms that affect throughput hikes can be amended in ‘n’, and backwards compatibility with both 2.4GHz 802.11g and 5GHz 802.11a must be aintained, with ‘n’ itself likely to run in both bands. However, it can include certain optional ‘greenfield’ modes that do not support mixed-mode systems including older Wi-Fi devices.
Among the approaches suggested in the study group phase were smart antennas, enhanced modulation, greater bandwidths and closed loop methods, which could be mixed and matched to boost PHY data rate close to 250Mbps and real throughput well over 100Mbps.
No clear directions have yet emerged though, except that any solutions that boost performance using more spectrum are unacceptable because of the shortage of spectrum in areas like the US, as are approaches that use channels wider than Wi-Fi’s normal 20MHz.
The Wi-Fi Alliance, the industry body that promotes and tests 802.11, has come up with the functional requirements and comparison tables to be used to evaluate 802.11n proposals, stressing that all devices should be suitable for home, enterprise and hotspot markets.
This is another in a string of examples of the Alliance taking a more and more active role in efining standards as well as just supporting them once they are ratified – it has already devised stopgap specifications for security and quality of service, for instance, to take the platform forward while users wait for true IEEE standards to be ratified. Many fear that this puts too much influence into the hands of the large vendors that dominate such bodies, but the growing role of the Alliance is also another sign that the market cannot live with the slow speed of traditional standards processes any longer, and will turn to other groups to give things a kickstart.
This is an issue that will certainly lurk behind all the politics and discussions that take place cross the various IEEE groups in Canada this week.
Like 802.11n, the 802.16 (WiMAX) group also expects its roadmaps to be speeded up, but there is no such optimism surrounding 802.15.3a, the personal area networking standard-inwaiting based on UltraWideBand. Although one of the two contenders to supply the standard, the Intel-Texas Instruments-led Multiband OFDM Alliance, will seek for a third time to have its technology adopted, few expect it to gain the 75% of votes required. Far more likely is that supporters of both technologies – the other being Motorola’s Direct Sequence CDMA – will pursue a de facto route and let the market decide.
Over at WiMAX, the 802.16 group expects to make significant progress on the mobile version of the specification, 802.16e, and has set a goal of publishing the standard within three to four months. However, Vancouver will also see a major gathering of the group devoted to the IEEE’s other broadband wireless technology, 802.20 or Mobile-Fi, which needs to accelerate its own agenda in order to keep its highly mobile, OFDM-based architecture on the map and ensure that its potential is not eclipsed by the less mobile, but increasingly competitive, 802.16e.
© Copyright 2004 Wireless Watch
Wireless Watch is published by Rethink Research, a London-based IT publishing and consulting firm. This weekly newsletter delivers in-depth analysis and market research of mobile and wireless for business. Subscription details are here.