[Techlunch] 802.11b question

[Wally]

some one asked the following question on another list 

any ideas?

Wally


---------------------------------------------------


Does anyone here happen to know much about the RF-level characteristics of
802.11b?

Specifically, if I have two cards, (a) and (b), both "capable" of 11
megabits per second, but because of signal levels are only actually
operating at 2 megabits/second, is the whole datarate slowed down to
2Mbps, or is the data rate actually 11Mbps but for only 20% of the time?

I ask this because I'm trying to understand (interpret) some power
readings.

If I pull a 10 megabyte file (80 megabits) from (a) to (b) and do so in 41
seconds, that's 1,951,219 bits per second. Since the cards are running at
2 Mbps, its (close enough) to 97% of the "possible" throughput.

If I monitor the output using a powermeter, I'm seeing +13dbm (ie, 20
milliwatts) output from a 100mW card.

Since we've got the card transmitting pretty well solidly (97%), I would
expect to be seeing 97% of the full output power (since this is an
averaging power meter), but I'm seeing roughly 20% of full (rated) power.

If the card were transmitting at 11 megabits/second, but only for 11/2 of
the time, the figures would sit close enough for me to be happy with.

But it doesn't make sense to me that the cards would fall back in this
regard - if the signal was sufficently poor that the cards had to drop
back in speed, then keeping the same symbol rate makes no sense, I'd
certainly expect it to slow the symbol rate down and therefore have to
transmit for the whole 97% of the time (see above time/data figures).

Anyone know any definitive information sources? I've not found anything so
far that actually states it one way or the other....