Direction finding for intermittent signal

Discuss KerberosSDR - 4x Coherent RTL-SDR
rtlsdrblog
Site Admin
Posts: 2386
Joined: Mon Nov 19, 2012 11:54 pm

Re: Direction finding for intermittent signal

Post by rtlsdrblog » Fri Aug 30, 2019 11:27 am

Not tested, but in the Android app turn off averaging, and put up the power threshold. If the signals are caught during the sampling time window they should work for localizing in the app.

But with intermittent signals there is obviously a lot less data, so it may take longer to get a good location fix.

253flo
Posts: 14
Joined: Mon Apr 22, 2019 6:20 pm

Re: Direction finding for intermittent signal

Post by 253flo » Sat Aug 31, 2019 2:50 pm

Df'ing intermittent sigs is by far the most interesting aspect and capability imo. Looking forward to taking it flying with me.

myke
Posts: 33
Joined: Wed Aug 28, 2019 7:03 pm

Re: Direction finding for intermittent signal

Post by myke » Sun Sep 01, 2019 3:38 am

Apologies if I missed this some place, and look forward to experimentation -- but looking at the Android App - Under Settings ->Filtering-> Minimum Required Power 0.0 Have any thoughts about the applicable measure of power ... dBm (such that negative value are expected ...ex: -117 dbm = .3 micro Volts) or is the value expect a specific range as you would find in a squelch setting such as a 0 to 10 value??

mousedude
Posts: 2
Joined: Fri Sep 27, 2019 7:06 pm

Re: Direction finding for intermittent signal

Post by mousedude » Fri Sep 27, 2019 7:51 pm

I already posted a related question before noticing this thread.

I'm a wildlife biologist, and this could be a game changer for tracking radio collared wildlife. The signals are very short duration pulses (20ms) at regular intervals (once every second or two).
If the system could be told the pulse interval and pulse width, and somehow pick the pulses out of the noise, that would allow automated radio tracking without using directional antennas.
Right now the only way to locate collared animals is using directional antennas. some researchers have set up arrays of multiple directional antennas for direction finding, but they were just using relative signal strength, rather than phase differences. The systems were huge and expensive.

to really be useful for locating a moving signal, you would need multiple stations that could each get simultaneous direction bearings on each pulse.

If I had a wish list of features:
the system could log the direction of each pulse, and multiple arrays could be used to triangulate the animal's location.
It might even be possible to incorporate some kind of mesh network to allow the receiving stations to relay their data to a base station where the data logging and triangulation could happen. If each station had a GPS and compass built in, you could just deploy several stations in a semi-random pattern across the landscape and they could each automatically determine their own location and orientation. The location/orientation could be included in the data being sent to the base station, and used in the triangulation calculations.

rtlsdrblog
Site Admin
Posts: 2386
Joined: Mon Nov 19, 2012 11:54 pm

Re: Direction finding for intermittent signal

Post by rtlsdrblog » Mon Sep 30, 2019 10:19 am

mousedude wrote:
Fri Sep 27, 2019 7:51 pm
I already posted a related question before noticing this thread.

I'm a wildlife biologist, and this could be a game changer for tracking radio collared wildlife. The signals are very short duration pulses (20ms) at regular intervals (once every second or two).
If the system could be told the pulse interval and pulse width, and somehow pick the pulses out of the noise, that would allow automated radio tracking without using directional antennas.
Right now the only way to locate collared animals is using directional antennas. some researchers have set up arrays of multiple directional antennas for direction finding, but they were just using relative signal strength, rather than phase differences. The systems were huge and expensive.

to really be useful for locating a moving signal, you would need multiple stations that could each get simultaneous direction bearings on each pulse.

If I had a wish list of features:
the system could log the direction of each pulse, and multiple arrays could be used to triangulate the animal's location.
It might even be possible to incorporate some kind of mesh network to allow the receiving stations to relay their data to a base station where the data logging and triangulation could happen. If each station had a GPS and compass built in, you could just deploy several stations in a semi-random pattern across the landscape and they could each automatically determine their own location and orientation. The location/orientation could be included in the data being sent to the base station, and used in the triangulation calculations.
Good ideas. We do intend to move in the direction of supporting networked units in the future.

There is this $40 program http://www.musther.net/RDFMapper/, made for another hardware radio + doppler based RDF system which already supports networked bearings. I think if we can emulate the COM port data it needs it would work.

I'm not sure i'd trust a magnetometer to be accurate enough. Maybe just use a standard compass to point the antennas North when deploying.

Still need to work on solving the intermittent signal problem first though.

rtlsdrblog
Site Admin
Posts: 2386
Joined: Mon Nov 19, 2012 11:54 pm

Re: Direction finding for intermittent signal

Post by rtlsdrblog » Wed Oct 02, 2019 1:24 am

Just confirming that intermittant signal detection should be possible, and it's going to be worked on in future updates.

g4yar
Posts: 7
Joined: Thu Sep 19, 2019 9:52 pm

Re: Direction finding for intermittent signal

Post by g4yar » Thu Oct 03, 2019 8:40 pm

Sounds excellent, I can’t wait for the networking possibility and a decent setting for intermittent signals. In the meantime is there anywhere that explains confidence and power values?
Thanks
PAUL

Benjamin73
Posts: 1
Joined: Fri Aug 23, 2019 6:21 pm

Re: Direction finding for intermittent signal

Post by Benjamin73 » Sun Oct 27, 2019 3:00 pm

Perhaps could we use the "Minimum required power" for intermittent signals such a DMR beacon? Has anybody tried playing with this parameter?

ckoval7
Posts: 11
Joined: Fri Sep 27, 2019 8:54 pm

Re: Direction finding for intermittent signal

Post by ckoval7 » Mon Oct 28, 2019 10:39 am

I've had plenty of success DFing signals that stay active for at least 3 seconds. For example, POCSAG. If a signal is any shorter than that, you're likely to miss it. I can't find much information on DMR beacons, other than someone on a forum post saying they're configured to be active active for 4320mS on a couple popular networks, which should be long enough to get good bearings off of. Setting "Minimum Required Power" between 5 and 10 gets rid of most of the bad LOBs.
The "power" value referenced everywhere in the KSDR software suite would be more accurately described as SNR (Signal to Noise ratio.)

rtlsdrblog
Site Admin
Posts: 2386
Joined: Mon Nov 19, 2012 11:54 pm

Re: Direction finding for intermittent signal

Post by rtlsdrblog » Tue Oct 29, 2019 10:42 am

The Power and Confidence values aren't calibrated to anything at the moment, so you just need to experiment with values that give good results for your setup.

Pulsed signals down to ~1s or even less should still give you good results.

If anyone has a good IQ recording of shorter pulsed signals e.g. from animal trackers, please email them to me [email protected], so we can test better squelching solutions.

Post Reply