May 12, 2003

Yes, Cory, Here the Celestial Harmonies Are Strong...

Scott Rosenberg reports his thoughts on watching two wireless computing gurus trying to figure out where to sit in a conference room.

Scott Rosenberg's Links & Comment: ...At one point at the start of a talk I saw Cory Doctorow and Glenn Fleishman wandering the hall, laptops held open at waist level, moving intently, deliberately, up aisles and down rows. I understood on an intellectual level that their orbit somehow involved a hunt for good 802.11b reception; but what my eyes took in seemed more seance-like, a wireless ritual. It was as though they were scouting for the geek music of the spheres. I hope they found it....

What's the wavelength of 802.11b signals anyway? A centimeter? What does that mean for the likely size of dead zones given whatever propensity they have to reflect off stuff?

Sigh. I'd know a lot of things I don't know had I been an engineering rather than a social studies major. (But I wouldn't be able to make fluid translations from the Trotskyist.)

Posted by DeLong at May 12, 2003 09:04 PM | TrackBack


The wavelength question is easy: 12.5 cm. As for the size and spacing of the dead zones: if they are due to interference of the waves, you're right to assume that they are usually on the scale of the wavelength. So they should just sit down anywhere and scoot their laptops around a bit to "tune them in".

There may be other issues involved, though: whatever is transmitting the signal (I'm not an engineer either - just a physicist who can speak on general principles) won't uniformly cover the whole room, so some regions may be "brighter" and some "dimmer" just due to geometry and reflection. A dead spot in an auditorium isn't so much about wave interference, but rather a spot where only the direct sound is arriving and little or no reflected sound.

So maybe they knew what they were doing after all.

Posted by: Ben Vollmayr-Lee on May 12, 2003 09:26 PM

I'm reminded of Robert Asprin's Myth Adventures books, where magic energy follows lines of latitude and longitude (or something similar); the closer one was to the line, the more power was available.

Posted by: Jon H on May 12, 2003 09:47 PM

What your realy looking for when you do that is where the hub and it's antenna are hidden, since they are often behind the wall in some utility closet nearby. My father gave me assorted wise advice one of which was avoid antenna design, but... in this case it's the antenna more than the wave length me thinks. The Wifi hub's antenna typically creates a shadow of signal that's donut shaped. A high gain antenna will create a flat wide donut, an lousy little antenna will create a round one of much lower radius with some of it's signal getting wasted warming the cloud cover.

Of course how you orient your laptop and it's antenna can effect things as well.

Sometimes the site has a directional antenna and then the signal is more beam shaped and there maybe a real advantage to getting into the middle of that beam.

The real rub is that after you sit down and get your charger plugged in, your coffee, and your VPN etc. setup some extra clever dude goes over and adjusts the antenna on the hub, the donut shifts, and you need to move again. Or worse yet the seminar starts and you have to listen to the speaker!

Posted by: Ben on May 13, 2003 05:30 AM

IIRC, in the comments to his post, somebody said that they were actually hunting for somebody's computer. Supposedly with Windows XP, when joining a conference, it's easy to have your computer declare that it is the host. This would then cause problems.

Posted by: Barry on May 13, 2003 06:09 AM

Exactly. And then people associate with the false base station and can't get on the network. We don't have a good way to find these. One tactic is to flood the host with packets and then walk around looking for someone who's 802.11b card is flashing like crazy.

Posted by: Eric Rescorla on May 13, 2003 08:12 AM

Yes, that's right -- it appears that they *were* hunting indeed, but not for good reception; rather for the hapless person who'd accidently created a rogue ad hoc network and thus screwed up access for everyone else... Glad the observation proved evocative anyway.

Posted by: Scott Rosenberg on May 13, 2003 09:44 AM

How to identify a radar engineer: it's the person who calls the rotating dingus in your microwave oven a 'mode breaker' rather than a 'fan'.

Posted by: Matt on May 13, 2003 09:55 AM
Post a comment