I searched and found some old posts from people asking about triangulation with HopeRF radios. But I didn't see any results posted. So I threw together some bases and a tracker with a display, all using 433 MHz RFM69W modules with quarter wave monopole antennas, to see what kind of results I could get. I didn't document the results because initial informal testing showed that there was too much variance in readings to be useful - likely due to obstacles, clutter, and antenna orientation.
Then I placed two of the same radios with quarter-wave monopole antenna about 0.5 km apart at about eight feet high on the edge of an open field. I then walked a similarly-configured receiver between them and away from them in a somewhat random pattern. I found that I was not able to clearly receive from the distant radio when I was on a line between the two and more than about a third of the way to one of them. This may be due to the fence and vegetation on the fence line between the two. When more distant from the two (out in the field), I noted a very rough but approximately linear relation between distance from each base radio though it varied quite a bit depending on objects between my receiver and a given base (such as a light pole or passing vehicle, maybe 15) and it at times varied by as much 10 at times for no apparent reason.
Note that this is all using the reported RSSI from the RFM69 library (radio.readRSSI() ) with the default modulation and rate. My code notes the RSSI of a ping reply for a given base promptly after receive. I can probably get significantly more range by lowering the rate.
No consideration was given to power saving, neither sleeping nor varying transmit power.
My preliminary conclusion is that these radios don't have enough sensitivity to give a good correlation between signal strength and distance even in a line of sight scenario. Objects in the way (cars, houses, etc) only add to the uncertainty. It will likely be impractical to have effective triangulation in a real world environment (urban or suburban neighborhood block) with only three or four fixed bases.
I want to try a few other things: using 915 MHz radios, reducing transmission rate, placing the bases at higher elevations, and a more complicated system involving more bases and varying transmission power to allow a wider range of data for interpolation.
All this to figure out where my cat goes off for half a day at a time.
Hi Jordan,
What you experienced is typical of radios in general.
Try nulling against the signal. There are several ways to do it.
It is not very difficult to make a 2 element Yagi.
Yagis work pretty well when properly made.
It also works to "shield" the antenna on one side.
For example, I have used a trash-can lid for direction finding on 147-MHz.
At 915-MHz, a gallon size can should work well.
Searching for "Direction Finding Antennas" should yeild lots of results.
Good Luck, John AE5HQ
What if the Moteino your cat carried had a GPS receiver?
Quote from: damonb on August 06, 2016, 09:38:46 PM
What if the Moteino your cat carried had a GPS receiver?
The places cats go often won't get reception. Also it won't have anywhere near the battery life keeping that thing cranked up.
Hi Jordan,
Even though Moteinos work well for their intended purpose,
they are not optimized for direction-finding.
I believe that you could achieve more consistent received signal strength indication
by reducing the capacitive effects of your body and any surrounding near-field objects.
If you are still experimenting, try this:
Place a 'large' ground-plane under the transmitter, such as a trash-can lid.
Keep it positioned above your body while you do your field test again.
I believe that by shielding your body from the antenna, the ground-plane
will provide a more stable transmitted signal level.
Again, as I mentioned previously, nulling a received signal is a
much more sensitive and reliable indicator of direction compared to
using the received signal strength as an indicator of distance.
Let us know how it goes.
Good Luck, John AE5HQ
Quote from: john4444 on August 06, 2016, 11:27:32 PM
Hi Jordan,
Even though Moteinos work well for their intended purpose,
they are not optimized for direction-finding.
I believe that you could achieve more consistent received signal strength indication
by reducing the capacitive effects of your body and any surrounding near-field objects.
If you are still experimenting, try this:
Place a 'large' ground-plane under the transmitter, such as a trash-can lid.
Keep it positioned above your body while you do your field test again.
I believe that by shielding your body from the antenna, the ground-plane
will provide a more stable transmitted signal level.
Again, as I mentioned previously, nulling a received signal is a
much more sensitive and reliable indicator of direction compared to
using the received signal strength as an indicator of distance.
Let us know how it goes.
Good Luck, John AE5HQ
I think your enter key is broken. And I'm jordan, not john.
Yep, I'm John, directing this comment to Jordan.
Was my suggestion of any help for you?