On 17 Apr 2004 17:36:48 -0700,
(E-Mail Removed) (Doiku) wrote:
>This kind of long-range WLAN stuff here in Japan is still much in its
>infancy.
There is a logarithmic relationship between the line-of-sight range
and the antenna amplification.
10 dB gain means that the range is multiplied by 10 (provided line og
sight between two identical directional antennas)
13 dB gain: range x 20
16 dB gain: range x 40
20 dB gain: range x 100
The calculations are, however, complicated by the fact that antenna
gain is measured in two different (often unspecified) ways: dBd or dBi
(the difference is between 2 and 3 dBs beteen the two as far as I
remember.)
Another problem is that at wi-fi frequencies (2.4 GHz) the antenuation
loss in the coax antenna feeder is MUCH higher than with e.g.
television cables.
If you use ordinary inexpensive RG-58 50 Ohm cable the loss is about 1
dB per meter.
LMR-400 cable has a loss of about 0.2 dB per meter but is more
expensive and difficult to find (if you build the antennas yourself.)
I would guess you need 2 unidirectional antennas with a gain about
15-18 dBi with short antenna feeder cables to get a really stable
connection. But the better the gain the less vulnarable to
interference it will be.
In some countries very high gain antennas could infringe on some radio
regulations ( I have no idea if that is the case in Japan but it could
well be a theoretical problem...)
Alex