title:

Faster than you need is not fast enough

By Scott Bradner

There is something in a standards organization that abhors a limit. The latest example is the work underway in the IEEE 802.11n project. There is a lot of wrangling to go yet but the goals of the project are impressive enough.

The High Throughput Study Group (802.11n) (http://grouper.ieee.org/groups/802/11/Reports/tgn_update.htm) was authorized by the IEEE about a year ago to define a new version of the wildly popular 802.11 Wireless LAN standard that could run at twice (or more) the speed of the 52 Mbps 802.11a/g standards. (The formation document for the study group (http://www.ieee802.org/11/Documents/DocumentHolder/2-798.zip) and the project criteria (http://www.ieee802.org/11/Documents/DocumentHolder/2-799.zip) are available on the IEEE web site.) The deadline for presentations to explain individual proposals for technology to meet the criteria were due mid August. Last June more than 60 people had indicated that they intended to make presentations about full or partial proposals but I've not yet seen how many actually submitted presentations.

Two groups seem to have decided that PR campaigns might help their chances when it comes to getting their proposals adopted. The first group, which is calling its proposal "TGn Sync," includes Agere Systems, Intel, Sony, Nokia, Philips and Atheros. There is a whitepaper on the TGnSync proposal at http://www.deviceforge.com/articles/AT5096801417.html. The other group calls its proposal "WWiSE" which stands for "WorldWide Spectrum Efficiency". The WWiSE group includes Texas Instruments, Broadcom, Conexant, STMicro, Airgo and Bermai and has a website at http://www.wwise.org/. The name of the second proposal alludes to one of the differences between them, the WWiSE proposal uses 20 MHz channels, which are supported in all countries and the TGn Sync proposal uses 40 MHz channels, which are not supported in Japan but can back off to 20 MHz channels. Both proposals include modes that are compatible with 802.11a/b/g and other modes that support 500 Mbps or faster data rates. One hopes that these two groups can work together, along with some subset of the many other proposers, to come up with a single standard within a reasonable period of time.

One interesting tact that the WWiSE group is taking is to offer, with some specific conditions, to license the technology in their proposal royalty-free to companies implementing the resulting IEEE standard (assuming their proposal becomes the standard). This may look better than it actually is because these days it's unlikely that the members of any particular group will own or control all of the patents that lawyers somewhere will assert apply to a standard. Just to make the intellectual property rights games more interesting, there is no reason to think that all the companies that may decide to assert patent rights in the future are currently participating in the IEEE discussions where they might have to disclose such patents or patent applications.

I'm sure that bandwidth-wasting applications will be developed (like the memory-wasting applications so popular for PCs these days) that will make use of the data rates that 802.11n will enable but, in the meantime, I can get the same vicarious thrill with wireless as I do now with my gigabit Ethernet-equipped laptop plugged into my home gigabit Ethernet LAN and connected to my rather non-gigabit cable modem.

I'm sure there are theoretical and practical limits to the data rates that can be teased out of this type of technology but, using history as a guide, I would not want to bet on what the standards 10 years from now will include.

disclaimer: Part of Harvard's job is the testing of perceived theoretical and practical limits on all sorts of things but I did not ask about this particular one thus the above opinion is mine.