Diary
44
Tests
on Hog Island
Part
1: Videoconferencing From Broadwater Tower
In December
2001, just to show that it was possible, we attempted a bit of
videoconferencing from Broadwater Tower to the mainland, using
900 MHz WiLan radios with 2 Mb bandwidth. Computer hardware was
one laptop at Broadwater, operated by Dr. John Porter, and a desktop
computer at the LTER lab in Oyster. Both machines ran NetMeeting
under MicroSoft Windows. There being no intra-island wireless
link at the time, Dr. Porter ran a 40-meter Ethernet cable from
the top of Broadwater Tower to his laptop, which was equipped
with a small digital camera.
John Porter, conferencing with people 14 miles
away
Screen Shot, with Randy Carlson foreground, John
Porter inset
Excerpts
from that video conference (as we experienced it at the lab in
Oyster) can be viewed by clicking on:
http://wireless.oldcolo.com/biology/Progress2001/video2001/oysterclips/oyster.html
Part
2: Link Test From Broadwater to Machipongo
On our
next visit in February of 2002 we tested a WiLan link between
Broadwater Tower and Machipongo Station, about 6 miles from south
to north.
Hog Island
Machipongo
Station is an old Coast Guard station that was used for various
rescue and reconnaissance tasks in the 1940's. Like much of Hog
Island, it is now the property of the Nature Conservancy. The
VCR LTER has tide measurement instrumentation located within the
Machipongo station, plus a meteorological station a few hundred
meters away.
Machipongo Station
The rather
unsurprising conclusion from testing WiLan radios (the same ones
we had used between Broadwater Tower and the Oyster lab) is that
if a 14 mile link works fine, a 6 mile link can be expected to
work even better.
At this
point it was necessary to decide how to set up two hops -- Oyster
lab to Broadwater, and Broadwater to Machipongo. We considered
several approaches.
- Use
the existing radio at Broadwater as a relay. This would be a
relatively inexpensive solution in that the whole job could
be done with a total of three radios -- which, happily, we had
already purchased. But that approach presented several disadvantages
as well. For one thing, the relay radio at Broadwater would
need to become the base station, from which all radio signal
diagnostics would be performed, requiring troubleshooters to
go 60 feet up a rusty tower. Also, this would cut our beloved
2 Mbps bandwidth by half (as it would be shared by to receiving
radios). Finally, in order to point in two directions from one
radio, it would be necessary to split the signal between two
antennas -- reducing antenna gain for each by 3 dB, including
the somewhat marginal link back to Oyster.
- Use
two WiLan radios at Broadwater, relaying from one to the other
via Ethernet. This would solve both the bandwidth problem and
the split antenna problem mentioned above. However, it would
also require the purchase of a fourth WiLan radio. Worse, it
would mean a doubling of power consumption, as the WiLans are
constantly transmitting, and drawing lots of power to do it.
- Use
the 802.11b wireless cloud as a relay. Since Broadwater Tower
was going to be set up as one of two 802.11b access points for
use on the island (the WiLan radios are just for the backhaul
to the lab), we decided to use that network as the "backhaul"
for Machipongo station.
In a meeting
with Dr. John Porter, Co-PI for the VCR LTER site, we chose the
third option to relay from Broadwater to Machipongo. This will
be discussed in more detail in a later diary.
Bonus
Link
Click here for a web page
(courtesy University of Virginia) with images from Broadwater
Tower's web cam.
-Tom Williams
Special Assistant
NSF Wireless Field Tests
Previous
Next
