While IT managers are debating the merits of the two 54Mbit/s Wi-Fi standards, 802.11a and 802.11b, there is even more confusion in store. Hardware vendors have rightly guessed that users have an insatiable desire for speed, and lined up an array of proprietary -- and incompatible -- technologies to make Wi-Fi go even faster.
The market for boosted wireless LANs, like the market for Wi-Fi itself, is focussed on 802.11g, the fast wireless standard that sticks to 2.4GHz, rather than 802.11a, the 5GHz specification. The reason for this is that much more 802.11g kit is being sold than 802.11a, and it is being sold to a market sector where add-on features such as speed boosts can be a big advantage for the vendor.
The technology is coming from wireless chip vendors. The two main options are Super G from Atheros, in D-Link and Netgear products, and Linksys which uses Afterburner from Broadcom.
But who is using them, how much trust should be put in them, and what are the prospects for a real fast Ethernet standard?
Too fast for the home, too risky for the office
802.11g is coming into use first in the home and in small offices, and in these environments, compatibility and long term stability aren't such an issue. Users probably only have one or two devices connecting to the network, and will not be overly concerned if their technology only works with certain client cards. It probably won't bother them if their particular go-faster 802.11g product disrupts rival 802.11g and 802.11b equipment (as has been reported). They can use whatever equipment they like. And with only one access point on the site, it's no big deal if a better technology comes out a year down the line, or if a final standard supersedes what they have. They can junk one access point.
Given that kind of user, the vendors are going all out with any speed boost technology that comes up. The competition is cut-throat, and vendors will use any short-term advantage to sell products.
The situation is more than a little ironic, because the .11g boosting technology is coming into play in the very environment where it is of least use. Homes and small offices are usually on the end of an ADSL connection, which gives a maximum of 2 Mbit/s and more likely about 500 kbit/s. The Internet connection is the bottleneck, and the speed boost will not be seen except in the comparatively rare instance of moving files from machine to machine in the home or office.
"The enterprise does not use pre-standard or turbo mode Wi-Fi," says Frank Ferro, Wi-Fi director at Agere, which recently came out with its own speed boost, Turbo G.
The best advice for IT managers is probably to ignore technology that boosts 802.11g, especially if your network involves multiple access points, and especially if you support a mixture of devices.
If you decide that you can build an enterprise network out of commodity consumer access points (and it is arguable that the technical work involved is worth the price savings), then you may well find you get a boosted .11g system by default, in which case you may want to turn it on and try it.
Alternatively, if you have a specialist area where faster speed would be a boon, where security is not much of a problem (so commodity access points don't worry you) and where you control the devices that attach, you may want to try some boosted g.
A follow-up article will look at the specific contenders for boosting 802.11g
Standards are a way off
The boosting technologies come from the makers of the basic wireless LAN chips in the kit, not from the system builders. Many vendors are using the same underlying technology -- for instance, both NetGear and D-Link use the Atheros Super G chipset which goes up to 108 Mbit/s.
But don't expect that to guarantee interoperability between, kit from NetGear and D-Link, as there are no interoperability standards covering this area. The Wi-Fi Alliance could be expected to weigh in, but as it is sponsored by those same vendors who are slugging it out in the market, its influence is limited.
So far the Alliance is maintaining a studied silence on go-fasters for 802.11g; products are certified at the .11g level, and any other features are ignored. There are those who feel this is a weak stance, as the boosted 802.11g signals are reported to disrupt transmissions of 802.11b and standard 802.11g networks. Why should a device that will spoil Wi-Fi compliant transmissions have a Wi-Fi certification sticker? Because to refuse it would go beyond the Alliance's political power.
However, at CeBIT, the Alliance spoke out in favour of faster Wi-Fi generally, putting its weight behind the only real prospect for a standard, the 802.11n group commissioned by the IEEE. However, this group is still at a very early stage. No standard is expected till 2005, and the group has not yet even decided what kind of technology it should use (it may use something like clever antenna technology such as MIMO). It is also quite likely that it will operate in the 802.11a frequency of 5GHz, in which case any prospect of standards in the 2.4GHz frequency where 802.11g and its variants operates will be up to market dominance or agreements between vendors outside formal standards bodies.
"802.11n will be here in 2006," says Ferro. "Right now, the group is looking at both 2.4GHz and 5GHz -- the proposal may cover both." Although the final standard won't be completed for a while, technology submissions are due in June 2004, and a basic physical layer data rate of around 100 Mbit/s will be the baseline.