Menu
Menu
Aggregating air: Toward optimizing wireless

Aggregating air: Toward optimizing wireless

While we are barely into the second quarter of the year, it is already reasonable to label 2003 the year of enterprise wireless

While we are barely into the second quarter of the year, it is already reasonable to label 2003 the year of enterprise wireless. Aruba Wireless Networks Inc., Trapeze Networks Inc. - and now Engim Inc. - are taking a distinctly enterprise look at harnessing wildly popular wireless. For its part, Engim is focusing on the most constrained resource: the thin air itself. A chipmaker coming out of stealth mode, the company says it has solved a lot of problems associated with delivering enterprise-class performance - or "wired experience" for corporate wireless LAN (WLAN) users. Ironically, the company has solved problems that, I'd bet, most enterprise network managers aren't aware they even have. Of course, that is part of the problem too.

The technical part of the problem is that, in a world where vendors are pushing dedicated, billion-bit-per-second desktop connections, the ubiquitous wireless standard, IEEE 802.11b, functions more like a 10M bit/sec Ethernet hub. It provides shared rather than dedicated bandwidth, and like the bad old days, high use by one user can translate into poor response time for other users. And that is with ideal conditions.

Interference from building materials, signal degradation caused by distance from the access point - even your colleague using the microwave - could cause users to drop to about 1M bit/sec communication with the access point.

So why no howls from users? Well, in many places, deployment is in the early stages, so performance problems have not yet manifested. As with cell phones, the benefits of mobility far outweigh problems with speed or quality. And if you complain, your IT guy might grab your wireless network interface card and shove a cable in your face.

There is a human aspect to the problem as well. Precious few of us old-line data networkers are radio frequency experts. As fast as you can say "Orthogonal Frequency Division Multiplexing" (OFDM), we're heading for the hills.

Today's plug-and-play wireless is about as complicated and fast-moving a technology as we've seen in a while. I always judge complexity relative to ATM - the gold standard of complexity. 802.11 wireless seems to have it beat - simultaneously shipping three "standard" flavors of the technology with more dot-eleven subcommittees seeming to spawn daily. And because our users are not complaining, why dig deeper when there are so many other things to do?

Current generation chipsets simply aren't designed to optimize the wireless environment. In the most basic sense, available bandwidth is not harnessed effectively. An 802.11b single-radio access point has three channels, but will use just one. This access point might be communicating with nearby users at multiple megabits per second while communicating with more distant users at less than 1M bit/sec. Given the round-robin approach used by access points when servicing clients, "fast" clients could end up waiting for slow clients to finish communication. Thus the presence of "slow" clients could degrade performance for all users of the access point.

The Engim approach is to "aggregate air" (my term, not theirs) by utilizing multiple channels simultaneously and optimizing other aspects of the radio frequency environment. When you apply these principals, say, to 802.11a, which offers eight 54M bit/sec channels, you get some impressive possibilities.

(By the way, if you feel compelled to learn about OFDM, you'll be happy to know that PaloWireless Pty. Ltd. has an OFDM Resource Center at www.palowireless.com/ofdm/ tutorials.asp.)

Tolly is president of The Tolly Group, a strategic consulting and independent testing company in Manasquan, N.J. He can be reached at ktolly@tolly.com. -- Network World (US)

Join the CIO New Zealand group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.

Join the newsletter!

Error: Please check your email address.
Show Comments