Uptake of wireless networking slowed by shortcomings in key technology

While wireless networking has taken off rapidly in the consumer space, enterprises have remained cautious because of three shortcomings of the key technology, IEEE 802.11x or Wi-Fi, which are all being addressed over the coming months. The three areas have been uncertain security, lack of quality of service standards (necessary for the increasingly interesting voice over Lan application), and limited real world data rates. Now, an upcoming extension to the standard, 802.11n, will support rates over 100Mbps, while a robust security technology, 802.11i, has finally been ratified, and first QoS moves are being taken.

FAST WI-FI

A battle has commenced at the IEEE to define the technology for the upcoming 802.11n standard for 100Mbps-plus Wi-Fi. Two main groups are submitting proposals, and are already engaged in a publicity war that includes technical claims and counter-claims, and promises over free licensing, that bring to mind the past year's UWB tussles. The danger is that, if the 802.11n process deadlocks too, the leading vendors will go ahead with their various approaches anyway, fragmenting the standard and worsening the non-interoperability dangers already raised by suppliers' 'standards-plus' extensions to vanilla 802.11x.

The first group is WWise (World Wide Spectrum Efficiency), which includes Texas Instruments, Broadcom, Conexant, STMicro, Airgo, and Bermai. Their proposal rests on using four MIMO smart antennas with conventional Wi-Fi channel widths of 20MHz, to achieve speeds of 135Mbps in standard mode, and up to 540Mbps peak.

The main rival proposal is from a group called TGnSynch, led by Intel, which flavors a combination of two MIMO smart antennas and wide (40MHz) channels to create a maximum throughput of 250Mbps (175Mbps of it usable at one time), and which it claims could scale to 600Mbps over time.

While WWise is dominated by WLan chipmakers, TGnSynch includes consumer electronics and mobile handset giants and so may be expected to carry greater weight in the wider market and in terms of OEM uptake. The group includes Agere Systems, Atheros, Cisco, Intel, Marvell Semiconductor, Nokia, Nortel Networks, Philips, Samsung, Sanyo Electric, Sony and Toshiba.

But WWise claims its rival's approach reduces the number of non-overlapping channels, negating some of the advantages of the current 802.11a standard, and will not be usable under Japanese spectrum rules. TGnSynch retorts that its proposed adaptive radio techniques will get round variations in regulations and will future proof the standard against changes and expansion in unlicensed spectrum availability in the future.

WWise has also announced, in a major publicity coup, that, if its proposal is accepted, member companies would license any patents necessary to implement the proposal to any company royalty-free.

The first draft of the 802.11n specification is scheduled to be completed in mid-2005, with its final ratification expected in late 2006 to early 2007.

Some analysts are hopeful that, since the two groups' technologies are nor vastly different, accommodation will be reached easily. However, as past experience shows, the fight will be over control and influence rather than technology per se. We fear 802.11n could turn into another UWB (ultraWideBand) experience (this standard is split).

Another worry about the standards process is the increasing influence of vendors rather than global bodies. The issue was highlighted last month when two important wireless focused meetings were held. One was the IEEE gathering in Berlin, to consider 802.11n and other crucial matters. The other was the Intel Developer Forum (IDF) in San Francisco where two technologies that pre-empt IEEE standards were paraded. There is increasing suspicion that the latter, a single vendor event that was once the preserve of the teccies, had more influence on the wireless future than the former.