Drone pilot tracking

[DroneTracker12]

Hey all,

I'm in my final year of university in computer science and wanted to look at pilot tracking.

I'm hoping to focus on only 2.4GHz bands due to my budget, using machine learning scanning 2.4GHz signals you can differentiate between drone communications and other devices.

Once you've found the drones communications would you be able to track the source of the controller going off the signal strength?

 

[Cafguy]

You are probably vastly more educated in this than I am BUT have you looked at some of the code they use for those RID apps I believe that indeed they can see where the pilot is in relation to the drone but not 100% sure. I did an experiment and built an RID "spoofer" using Arduino and used one of those apps to test it but have not used it since that's pretty much my expertise when it comes drone Tracking.

 

[Yorkshire_Pud]

Hey all,

I'm in my final year of university in computer science and wanted to look at pilot tracking.

I'm hoping to focus on only 2.4GHz bands due to my budget, using machine learning scanning 2.4GHz signals you can differentiate between drone communications and other devices.

Once you've found the drones communications would you be able to track the source of the controller going off the signal strength?

Unless it is the method of tracking that is 'new', have you checked the idea with your supervisor?

I ask because I think such things already exist.
It seems to be one of the aims of RID and prior to that DJI had some sort of scanner that could, I think, do the same thing. I have forgotten its name.

 

[DroneTracker12]

Unless it is the method of tracking that is 'new', have you checked the idea with your supervisor?

I ask because I think such things already exist.
It seems to be one of the aims of RID and prior to that DJI had some sort of scanner that could, I think, do the same thing. I have forgotten its name.

I've looked further into it and it looks like I'd need to use an frequency scanner, apply the ML to distinguish the communication line between the two nodes. Then using an RF finder. This is way out of my budget and my ability as I don't have any prior knowledge on RF.

Does anyone have any ideas for me?

 

[Felix le Chat]

Hey all,

I'm in my final year of university in computer science and wanted to look at pilot tracking.

I'm hoping to focus on only 2.4GHz bands due to my budget, using machine learning scanning 2.4GHz signals you can differentiate between drone communications and other devices.

Once you've found the drones communications would you be able to track the source of the controller going off the signal strength?

Used to do a similar trick tracking CB signals in the late 70's (higher strength: closer to the transmitter) so, yes it would be possible, but hit-and-miss with accuracy using a dipole. You'd do better with a Yagi-Uda because they are highly directional. This technique has been used for over 80 years.
DJI AEROSCOPE already provides very accurate drone and controller tracking, scanning 2.4 and 5.8ghz for ocusync and lightbridge transmissions. This live transmission contains progressive GPS location for the drone as well as the controller. Individual Aeroscope receivers are deployed as a network to cover larger geographical areas.

 

[GFields]

Greetings from Birmingham Alabama USA, welcome to the forum! We look forward to hearing from you!

As Felix le Chat pointed out, I think this has already been done.

 

[HoozierDroneDaddy]

Unless it is the method of tracking that is 'new', have you checked the idea with your supervisor?

I ask because I think such things already exist.
It seems to be one of the aims of RID and prior to that DJI had some sort of scanner that could, I think, do the same thing. I have forgotten its name.

It was called Aeroscope

 

[DroneTracker12]

Used to do a similar trick tracking CB signals in the late 70's (higher strength: closer to the transmitter) so, yes it would be possible, but hit-and-miss with accuracy using a dipole. You'd do better with a Yagi-Uda because they are highly directional. This technique has been used for over 80 years.
DJI AEROSCOPE already provides very accurate drone and controller tracking, scanning 2.4 and 5.8ghz for ocusync and lightbridge transmissions. This live transmission contains progressive GPS location for the drone as well as the controller. Individual Aeroscope receivers are deployed as a network to cover larger geographical areas.

I've thought about it further, I'm also trying to go as cheap as possible. I'll be attempting to track cheap hobby drones only on the 2.4 frequency.

This is a question for anyone but with the dataset used here: DeapSECURE module 3: Machine Learning: Case Study 2: Drone RF Signal Classification

If you where close range, with an antenna and you where getting closer to the controller would the strength increase?

In that link it doesn't explain what tool the data was extracted from, does anyone know what that could possibly be?

 

[Felix le Chat]

I've thought about it further, I'm also trying to go as cheap as possible. I'll be attempting to track cheap hobby drones only on the 2.4 frequency.

This is a question for anyone but with the dataset used here: DeapSECURE module 3: Machine Learning: Case Study 2: Drone RF Signal Classification

If you where close range, with an antenna and you where getting closer to the controller would the strength increase?

In that link it doesn't explain what tool the data was extracted from, does anyone know what that could possibly be?

"...If you were close range with an antenna and you where getting closer to the controller would the strength increase?..."

Yes.

The techniques are already there. The only thing that has altered are the broadcast frequencies.

Single antenna detects strength of signal. Three mobile units with adjustable directional antennas at different localized geographical points gives triangulation... exactly what the German army did to detect and efficiently locate French resistance transmissions during the 1940's. The Brits did the same in tracking down active 5th Columnists and spies.

Go to Wikipedia, enter Bellini–Tosi direction finder in the search bar. Follow this down the rabbit hole.

Alternatively: if you want something a bit more 21st century: you could head over to Youtube and check out Mads Tech - he has an interesting article featuring a tear-down and analysis of a DJI Aeroscope unit. DJI claimed (for quite some time) that the UUID and RID transmissions broadcast from their drones since the P3(?) were encrypted, after a while they finally blew the gaff that they weren't... and still aren't.

With RID scanners popping up like mushrooms all over the app stores, if you're considering developing something for commercial purposes: with respect, I think you might be trying to re-invent the wheel, but as an exercise? an interesting project. Wartime history is always the best place to use as a datum point... that's where the greatest leaps forward in technological development always start. You might think that this is a bit simple - but the thing about simple is... it works.

 

[DroneTracker12]

"...If you were close range with an antenna and you where getting closer to the controller would the strength increase?..."

Yes.

The techniques are already there. The only thing that has altered are the broadcast frequencies.

Single antenna detects strength of signal. Three mobile units with adjustable directional antennas at different localized geographical points gives triangulation... exactly what the German army did to detect and efficiently locate French resistance transmissions during the 1940's. The Brits did the same in tracking down active 5th Columnists and spies.

Go to Wikipedia, enter Bellini–Tosi direction finder in the search bar. Follow this down the rabbit hole.

Alternatively: if you want something a bit more 21st century: you could head over to Youtube and check out Mads Tech - he has an interesting article featuring a tear-down and analysis of a DJI Aeroscope unit. DJI claimed (for quite some time) that the UUID and RID transmissions broadcast from their drones since the P3(?) were encrypted, after a while they finally blew the gaff that they weren't... and still aren't.

With RID scanners popping up like mushrooms all over the app stores, if you're considering developing something for commercial purposes: with respect, I think you might be trying to re-invent the wheel, but as an exercise? an interesting project. Wartime history is always the best place to use as a datum point... that's where the greatest leaps forward in technological development always start. You might think that this is a bit simple - but the thing about simple is... it works.

Thanks for the info, I was also considering a frequency analyser to find the signals and also show the distance of the pilot via the strength of the signal. The guide to this seems reliable: VMA Simple Spectrum Analyser - Download the latest version here!

Does anyone have any very extremely cheap drones which would work for this project?

Does anyone have any good sources for frequency analysers for only the 2.4GHz and surrounding bands for scanning RF? When I say cheap £80 maximum.

This seems to be getting a bit over my head, I may need to change my project due to my total lack of knowledge on RF.

 

[DroneTracker12]

I've looked into it further and it's out of my price range going for 2.4GHz drones, I may look at tracking walkie talkies at lower frequencies.