Information from JimThompson about the physical restrictions you run into when trying to do very long 802.11b links (from the NoCat MailingList).

I think I covered this over on the HostApMode list, but...

In a 802.11 DS system with an 11Mb/s rate, the timing of a transmission is as follows.

Each data packet in one direction is followed by an acknowledgement from the recipient in the opposite direction. The data packet typically lasts 1210µs, the ACK 106µs and there is a SIFS, (10µs delay) between data packet and the ACK. Information is typically streamed in a continuous manner: the next packet arrives 350µs after the end of the previous ACK. The information exchange is asymmetric: either the AP is sending data packets and the STA is sending acknowledgements, or the situation is reversed.

{{{ TX RX TX

_| || |_| |_

Now, 'uS' are not 'mS' (by 3 orders of magnitude!) So the range below (and in the subsequent message) are.. optimisitic.

One 'uS' is 1/1,000,000 of a second. (One 'ms' is 1/1000 of a second.) So, the fundamental limitation is SIFS+ACK, or, how far will RF go in 116uS (@ 11Mbps).

In practice, you'll play hell getting an 802.11b signal to go 21 miles due to fading issues, etc. So the NICs will be your friend, and reduce the 'speed' of the link, first to 5.5Mbps then 2Mbps, then 1Mbps. As the speed of the link goes down, the time in flight goes up, which accounts for all those links we hear about that are > 20 miles.


{{{> Something else I have been investigating is the rf absorbtion spectrum of > water. Water seems to start it's absorbtion bell curve at about 11Ghz, > peaking at 23Ghz. The absorbtion curve never fully tapers off, as oxygen's > (O2) bell curve begins to mount, peaking at 69Ghz. > > The loss due to humidity in the atmosphere peaks at ? db per kilometer. I > suspect our LOS prohibiter is carbon not water. Weather or not trees are > bags of water or bags of carbon is immaterial. Trees stop 802.llb cold. More > to follow... Spring brake is here so I got plent of time on my hands.}}}

Your suspicions are unfounded, but perhaps you'll enjoy reading the following during spring break.


MaximumDistance (last edited 2007-11-23 18:02:29 by localhost)