FINAL
VISIT TROUT LAKE LTER
Experimentation and modeling of wireless support for the environmental
research work at Trout Lake, Wisconsin has proven quite successful.
The Trout Lake Station, directed by Dr. Tim Kratz, carries out a
wide range of long term fresh water studies on and around the large
number of lakes of northern Wisconsin.
The bodies of water are not large, rarely exceeding one mile across,
but they are invariably surrounded by tall deciduous and pine trees
averaging 80 feet in height. It is a heavily wooded area. This characteristic
posed substantial challenges to the use of wireless, since the most
cost effective digital radios, operating in the FCC Part 15 bands
of 902-928Mhz, 2.4-2.483Ghz, and 5.8Ghz are limited in their tree
penetration ability by both FCC permitted power levels, and the
frequencies allocated. With one singular exception - a small hill
called Muskie Mountain - there are no hills or high points from
which to communicate over the trees.
Since the Research Station, its facilities and terminating point
for long line Internet connections, are on the southern edge of
Trout Lake itself, in the center of a study area involving up to
100 lakes, it was clear that the first task was to get a high point
from which to communicate.
Our NSF Wireless project contracted for the erection of a 120 foot,
guyed tower in 1998, the first year of the 3 year project, close
to a storage building which had wall power and a small room in which
radios could be housed. It was also within 100 feet of the office
facilities where the T-1 Internet link to the University of Wisconsin,
Madison terminated, and served all the devices in that building,
and, Ethernet wired, made the net available in all the housing cabins
where researchers stayed. The tower was designed, and sited to make
the least intrusive structure at the rural site. There was some
concern that it would resemble the very large telephone company
cell towers which can be visually extremely distracting and overpowering.
However the Trout Lake tower was very narrow at the base, and researchers
noted that it was hard to pick up against the sky when boats were
only a quarter mile off shore.
But it was critical to get a pair of antennas well above the trees
- we selected 40 feet higher than the surrounding tree tops - both
to reach over them, but also to present a lesser angle of entry
to water bodies through surrounding trees of the electromagnetic
energy.
Two antennas were placed on top - a 902-928mhz Omni antenna, with
8db of gain, and next to it, a 2.4Ghz Omni, also with 8db of gain.
Two 150 foot lengths of LMR600 RF cable - rated at 2db of loss per
hundred feet - was installed on the tower and ran about 35 feet
into the radio room.
This worked fine to both the Sparkling Lake Raft, sensors, and Data
Logger, and for the time it was in Trout Lake, the big Buoy.
The Sparkling Lake raft, sensors, and Freewave radio has worked
reliably and consistently. It was connected in the spring of 2002
to an NL100 operating off the NT Server of the Trout Lake Station,
which itself was connected to the T-1 Internet connection running
to Wisconsin University in Madison. Since then, the data out on
Sparkling lake has been fetched daily by custom scripts and moved
into Data Bases at Madison. But ALSO the Data Logger can be reached
from any Windows computer running PC208W v3.2 software over the
Internet. The big difference between the Sparkling Lake wireless
data connection and the ones in Alaska, was that solar/battery power
has worked year round in Northern Wisconsin so that the radio is
'always on' and data can be fetched at will from Sparkling Lake,
while only hourly from the 30 Alaska data loggers.
While
the custom processor and a Cisco 802.11b radio attached to it on
the Big Buoy worked back across Trout Lake (up to 2 miles) to the
radio antennas at Trout Lake for a summer, other problems required
the buoy to be pulled from the water for other repairs.
GREAT
SUCCESS WITH LITTLE BOUYS
We
had experimented earlier at Trout Lake and demonstrated to the LTER
staff, with World Wireless small radios as well as Freewaves connected
to data loggers on small buoys put out a quarter to half a mile.
While the buoys were out on the water, the researcher could do 'drive
by data collection' using a radio and laptop - usually powered from
a cigarette lighter in a car or truck - from the shore.
One research team headed by Paul Hanson, used that combination with
great success all summer of 2002. In this case a small buoy was
put out and anchored only for relatively short periods on different
lakes, which read carbon dioxide data from an advanced sensor system.
Then it was taken out and moved to other lakes.
Hanson reported that, while doing it the traditional way they could
cover only 7-8 lakes over a summer of time. With the drive by wireless
technique they were able to cover 30 lakes in one summer. An outstanding
use of wireless for data collection.
WATER
IN THE CABLES
A separate problem occurred late in 2001. The 150 feet run of LMR600
cable that connected the Freewave base radio to its 902-928mhz omni
antenna on top of the tower got water into it. And the system stopped
operating. It was not clear how it got into it, but liquid water
came out the bottom of the cable when the connector was detached,
so it is likely a bad seal at the antenna was the culprit. Underscoring
the extreme importance of sealing up all the orifices well at the
time cables are installed. Weather changes, freezing and thawing,
aging of tape all can cause cracks to develop. Whatever it was,
since the 120 foot tower required an expert climber to repair, we
had to retain the same company that put up the tower to go up and
first switch over the cable from the 2.4ghz antenna to the 915mhz
antenna, then switch connectors at the bottom to enable the Freewave
to continue operating. Disabling the second cable.
Then, in September of this year, we retained the same climber to
come out and totally replace the water logged LMR600 with a new
one - since once the inside of RF cables get wet there is no way
to dry them out. We required the climber - who is not an expert
in wireless - to use a product called 'Stuff' that is waterproof
but does not conduct electricity and can be squirted inside connectors,
adding to the seal provided by good electrical external taping.
This worked, of course, although at the added expense of two trips
up the tower, and purchase and replacement of 150 feet specialized
cable.
While this work was being done, we reconnected an Aironet 4800 802.11b
radio at the bottom, and made it possible to use 802.11b radios
for shorter range, high bandwidth - such as video - wireless.
The outdoor wired connections to wireless do not last forever, and
always represent a 'point of failure' in a wireless network.
OTHER WIRELESS ACTIVITIES
Once the Trout Lake LTER team gained some experience with the initial
experiments our NSF Wireless team deployed in 2000 and 2001, they
moved aggressively to expand their uses themselves.
Data loggers attached to Freewave radios have been deployed in a
specialized small water body called 'The Bog' several miles across
Trout Lake.
The data from Sparkling Lake has been ported into a web site, read
hourly:
http://144.92.62.239/SparklingData/profile2/
The 802.11b radio with its site-covering antenna on the tower is
going to be used to reach all the researcher 'cabins' on the site's
premises, so that researchers with laptops and 802.11b cards in
them can connect out to the Internet when they are doing their desk
work. Only some of the cabins hitherto had been Ethernet wired connected.
Video cameras will be set out on the lakes, to record other metrological
events. This is where the higher speeds of 802.11b radios will be
useful.
Experiments will take place trying to link the data from some costly
($1,000 or more) specialized sensors from which they need data from
many places.
Dr. Tim Kratz has applied to the FCC for an 'Experimental License'
under Part 5 of FCC rules to permit using higher power radios (or
higher - than 36dBi gain - radios) to reach better through the forests
into the lakes.
CONCLUSIONS
The wireless devices this project deployed, demonstrated, and which
worked, did what we hoped. First in supporting Trout Lake LTER's
field science. Secondly prove out what works and what does not.
And thirdly accomplished a 'technology transfer' to the Researchers
and their technical staffs and graduate students. Which they are
now expanding, reporting on in professional papers, and at conferences.
And they don't need our help any more.
