Sensorgnome update for Cornell Tag XCVR

Tagging
1

Hi folks,
Some people have been approached by David Winkler's group at Cornell
asking whether they can piggy-back a receiver for their solar-powered
tags onto a sensorgnome site.
This is now possible: one or more Cornell Tag XCVR USB radio dongles
can be plugged into spare USB ports on a sensorgnome, and detections
of Cornell coded tags will be recorded in the usual sensorgnome data
files. The XCVR dongle draws little additional power, so should not
affect SG operation.

A software update to permit this is here:

   https://sensorgnome.org/Software_for_Deploying_a_Sensorgnome/Release_History

and further details are here:

   https://sensorgnome.org/Experimental_Tags#section_1

Please direct any questions about this to one of the mailing lists.

J.

2

This is good news. I do have one question. My impression was that in order to use the Cornell tags, the SG needed to be built with a raspberry pi instead of a beagle bone. Is that still the case?

David Mathiason

3

Here are some platform options for detecting Cornell tags.
(a) is what works now; (b-d) will depend on user interest:

(a) working: Beaglebone-SG + Cornell Tag XCVR USB radio dongle

(b) TODO (easy): Pi-SG + Cornell Tag XCVR USB radio dongle

(c) TODO (hard): Pi-SG + rtlsdr dongle (needs a new processing plugin)

(d) TODO (hard): Beaglebone-SG + rtlsdr dongle (work of (c) plus easy port
                                 of rtlsdr support to Beaglebone-SG)

J.

4

John,

(c) suggests that your current RPi + rtlsdr hard work isn’t sufficient or needs a re-write?

And as I’ve acknowledged privately but now do so publicly, we have considerable interest in (c) and (d) and, to a lesser extent (b).

Thanks,
Adam

5

It's just that the Cornell tags emit a more complicated signal than
the Lotek tags; their XCVR radio dongle decodes that signal into
an ID code, but to use an rtlsdr for the same purpose would require
doing that decoding in a software plugin, as we do for the much
simpler Lotek tag signals.

6

Hi John,

I real love that there are coming more suppliers of tags. That’s real great! And lovely that you are working hard on it for the Motus database.

Earlier, I understood that this tags can also run on 150.100 mHz, what is great for Europe. But do I understand it well that this tags can’t detect by Funcube+ or other type of dongles? And if you have 6 antennas on your SG and like to detect Lotek and Cornell tags (if they have the same frequency range), do I need to put than two USB dongles on one antenna? And if yes: is putting two receiver dongles physically possible, without loss of detection range?

I’d like to hear of you!

René

7

Hi René

Earlier, I understood that this tags can also run on 150.100 mHz, what is
great for Europe.

The current Cornell tags operate at 433 MHz. Perhaps someone from
that group could comment on plans for other frequencies.

But do I understand it well that this tags can't detect by Funcube+
or other type of dongles?

There are three sampling rates available on the funcubedongle,
with these tradeoffs:

- 48 kHz - usual SG/motus setting; supports 6 or maybe 7 antennas;
   can't detect Cornell tags

- 96 kHz - should support 4 antennas; should detect Cornell tags, but
   has not been tested

- 192 kHz - only supports 3 antennas; can(*) detect Cornell tags.

(*) automatically extracting Cornell tag IDs from the signal would
require writing a plugin which, we haven't done. When testing, we
simply recorded raw signal from the funcube and examined it manually.

And if you have 6 antennas on your SG and like to detect Lotek and
Cornell tags (if they have the same frequency range), do I need to
put than two USB dongles on one antenna?

Yes.

And if yes: is putting two receiver dongles physically possible,
without loss of detection range?

A low-loss antenna splitter would send nearly 50% of the signal power to each
dongle, so each dongle's detection range would be ~ 70 % of that when
connected alone (i.e. splitting reduces detection range by ~ 30%)

J.