User's Manual
Rev B
27
after settling. Using a global noise threshold of 2, the gain settled to the desired
value of 56, with the following result:
10/11/89 15:40:08 151.450*64*110 151.138*96*97 151.149*2R*188
151.158*3R*109
10/11/89 15:41:09 151.450*64*92 151.138*96*91 151.149*2R*194
151.158*3R*140
10/11/89 15:42:01 151.450*64*90 151.138*96*113 151.149*2R*195
10/11/89 15:43:02 151.138*94*86 151.149*2R*195
10/11/89 15:44:20 151.450*64*90 151.138*96*98 151.149*2R*196
151.158*3R*127
10/11/89 15:45:47 151.450*64*80 151.138*96*80 151.149*2R*202
151.158*3R*116
10/11/89 15:46:13 151.149*2R*194 151.158*3R*114 151.450*64*100
151.138*96*104
There is a subtle improvement. We have now essentially optimized Event_Log's
noise filters.
Example 3: Automated Data Collection
You are studying the spawning behaviour of salmon on a major river system. You want to
find the most heavily frequented spawning areas and also to assess the effect of several man-
made structures, specifically three large hydroelectric dams, along major routes. To identify
spawning areas you plan to locate fixed data collection stations at critical branch nodes along the
main streams and, after examining the data from these stations, to use aircraft to search for the
upstream spawning beds. In the vicinity of the dams, however, you require more spatial
resolution, in particular around fish ladders and spillways. Your plan is to analyze relative
signal strength data from groups of local antennas placed at various entrances and exits, using
two receivers per dam.
One of the most significant tradeoffs you have to make is between overall sample size and
temporal resolution. For example, with 200 fish on 200 individual frequencies and with 6 local
antennas located around the entrance to a fish ladder the time required to look for each animal
once on each antenna is 1200(t + n) seconds, where t is the transmitter pulse interval and n is the
receiver processing time (typically about 200 msec.). For pulsed carrier transmitters operating
once per second this gives a round trip time of 24 minutes, which you judge to be barely
acceptable near an entrance but at an exit would translate to a fairly high probability of missed
data. Moreover, to meet the range requirements for the experiment, especially around the dams
where the water may be deep, your transmitted power limits the transmitter life expectancy to
about one month, which effectively disables the spawning study.
To cope with these conflicts you have decided to use coded transmitters. With a set of 25
codes you now only need to scan 8 frequencies and your cycle time becomes 48(t + n). With a
code interval of 5 seconds this translates to a cycle time of only 4 minutes, and allows the
transmitters to last more than 6 months.
Your system specification now includes six receivers for the dams plus another six which will
be located at branches in the river. Once the fish have passed the lower dams you may relocate
more receivers to other branches further upstream, or use them for tracking by boat or aircraft.
You will also need six ASP_8 antenna switching units, appropriate antennas and a quantity of
50-ohm (preferably low-loss) coaxial cable. You will be using Lotek's proprietary coded fish tags
and software version W16 (or a relative of it). You still have a logistical problem, however, in
that your stations are widely separated. Servicing your receivers, either for data retrieval or for
"tuning" of system performance is extremely labour intensive. Consequently you have decided