Selecting the right direction finding method

pommes

Newbie
Joined
Jan 31, 2025
Messages
3
Hi, Im new to direction finding and radio topic in general. I got into it 6 months ago because Ive realized how much we are surrounded by the electromagnetic waves and how much a signal can tell. So now im here doing my bachelor project about RF data visualization in VR/AR and sound.


Background
So far I learned SDRs and played a lot with RTL-SDR. I also developed a mobile spectrum sonifier based on Raspberry Pi that does a sweep scan, compares the data with the frequency allocation database of my country and turns it into sound. I also did my research on spectrum scanning and direction finding. And I know how to code, so custom software is not a problem.

Objective
My goal is to create a mobile device (e.g. based on Raspberry Pi+SDR) that would DF incoming peaks in its (instantaneous) bandwidth in real time, and then display the directions+data around the user using VR/AR. The size should be portable and the user should be able to walk with it. And I want to be able to scan the whole possible spectrum range supported by the SDR (somewhere from 70MHz to 6GHz). The DF will be done in a busy urban environment. And of course my budget is limited to max 700$ for the whole technical part.

Questions
Which DF method should I choose? What equipment do I need?

My thoughts on this
I already know that there is no perfect solution to this problem, but I tried to develop some concepts that would meet my expectations.


Pseudo-Doppler
HackRF+Opera Cake Switch with 4+ Antenna Array (Circular)
Pluto+ SDR+Switch with 4+ Antenna Array (Circular)
4+ RTLSDR working in sync
From my knowledge I could say that this is an easier solution for automatic DF, but it is not as good because of the multipath problem. Also, there is no need for coherent channels, so I do not need to spend my money on multichannel SDR or multiple SDRs.
Directional Antenna Array
SDR with directional Array (Linear/Matrix/Circular)
This would be a good solution since you can only look in one direction when wearing VR/AR. So mounting directional antennas directly on your headset would work for your field of view (or you could mount them around). The biggest problem is that wideband directional antennas are too big and expensive, but I've read about patch antenna arrays (that switch depending on the band), so maybe I could make custom PCBs with such an array and a switch (gets too complicated). Do you know of a directional antenna that is small enough? And is it mandatory that the channels are synchronized or can I just switch very fast between the antennas and lose some milliseconds?
Beamforming
4+ RTLSDR working in sync
Pluto+ SDR with 2 channel RX
The need to change the antenna spacing for different bands kills the mobility/portability. And I still do not fully understand the principle.
Correlative Interferometry
Pluto+ SDR with 2 channel RX (inaccurate)
2 Pluto+ SDR with 2 channel RX in sync (4 channels totally)
3+ RTL SDR working in sync
I think this is the most interesting method. The problem is that you need precise engineering and calibration to make it work. On the other hand, it's not affected by multipath as much as Doppler. So I've seen KrakenSDR and I know it uses this technique. It's great because you can find a lot of footage of it in action. And what I also found was this article by Epiq and Deepsig where they did PDoA with two antennas using ML. So I could imagine training a model for DF, but I do not have enough knowledge about the process behind it to actually create a dataset.


So for SDR I would prefer Pluto+ SDR, its quite affordable (150-200$), has IBW of 20MHz and covers a range from 70MHz to 6GHz. I would like to have the directional antenna array as it makes the most sense in my head. But if that's not possible, I can't decide between Pseudo-Doppler and Correlative Interferometry, it depends on technical requirements, information accessibility and price. Is there a commonly used method that I have missed?


Thanks for your time and help!
 

dlwtrunked

Member
Premium Subscriber
Joined
Dec 19, 2002
Messages
2,436
I have fancy stuff including even a KrankenRF, various Doppler, etc. but usually go simple and portable out of the car with a laptop. You should start this way rather than jump to more complicated at first step. You will learn a lot. Simple VHF/UHF DF loop and mount to a wood stick:
As it is a loop, you go for the nulls which are 90 deg off. Turn the SDR AGC off!!!! (If you cannot turn it off, get a different SDR.) And use you eyes (not your ears) on the SDR software to find the direction of the null. Point for dip down in signal strength. Get a good compass and know how to use it (identify something in the direction of the null, put down the loop, and then use the compass in that object correcting for magnetic declination. Every time, this has worked better than Yagis and arrays for me. People like to build Yagis, but i laugh as they have not even tried the better way. I have used this to find many unknown repeater locations. I use an AirSpy R2 which is fine. Note some of the fancier ways to DF will have trouble with digital signals. Again, start cheap and simple or you may end up with an expensive museum on unused. (***NEVER*** be a driver of a vehicle trying to DF--if you do, I can almost guarantee you will wreck you car no matter how careful--word from direct experience. )
 

pommes

Newbie
Joined
Jan 31, 2025
Messages
3
I have fancy stuff including even a KrankenRF, various Doppler, etc. but usually go simple and portable out of the car with a laptop. You should start this way rather than jump to more complicated at first step. You will learn a lot. Simple VHF/UHF DF loop and mount to a wood stick:
As it is a loop, you go for the nulls which are 90 deg off. Turn the SDR AGC off!!!! (If you cannot turn it off, get a different SDR.) And use you eyes (not your ears) on the SDR software to find the direction of the null. Point for dip down in signal strength. Get a good compass and know how to use it (identify something in the direction of the null, put down the loop, and then use the compass in that object correcting for magnetic declination. Every time, this has worked better than Yagis and arrays for me. People like to build Yagis, but i laugh as they have not even tried the better way. I have used this to find many unknown repeater locations. I use an AirSpy R2 which is fine. Note some of the fancier ways to DF will have trouble with digital signals. Again, start cheap and simple or you may end up with an expensive museum on unused. (***NEVER*** be a driver of a vehicle trying to DF--if you do, I can almost guarantee you will wreck you car no matter how careful--word from direct experience. )
Thank you very much. That's exactly what I thought, I lack practical experience. And such method would allow me to play before I would automate things. I found some videos about it and how to build such an antenna, so it should be easy. Have you ever worked with (very) small directional antennas?
 

dlwtrunked

Member
Premium Subscriber
Joined
Dec 19, 2002
Messages
2,436
Thank you very much. That's exactly what I thought, I lack practical experience. And such method would allow me to play before I would automate things. I found some videos about it and how to build such an antenna, so it should be easy. Have you ever worked with (very) small directional antennas?
I would buy rather than build at this stage to avoid building errors/flaws/disappointments. Those antennas are not just loops as they are also shielded, matching transformers, and you cannot easily build one. But they are relatively cheap. I always keep one of this on an 18 inch stick in the car and have located everything from US government repeaters to PLS (Project Life-Saver) bracelets with it. Slow, careful, no AGC and eyes not ears are the main points with one.

Regarding very small directional antenna, I am not clear of exact sort you mean. I have used Vivaldi in the GH z range but really just because I needed a antenna--not for its direcional properties. Ans if I knew something better for DF work than the loop and smaller, I would buy it if I saw it --like I do no have enough antennas :)
 

pommes

Newbie
Joined
Jan 31, 2025
Messages
3
I would buy rather than build at this stage to avoid building errors/flaws/disappointments. Those antennas are not just loops as they are also shielded, matching transformers, and you cannot easily build one. But they are relatively cheap. I always keep one of this on an 18 inch stick in the car and have located everything from US government repeaters to PLS (Project Life-Saver) bracelets with it. Slow, careful, no AGC and eyes not ears are the main points with one.

Regarding very small directional antenna, I am not clear of exact sort you mean. I have used Vivaldi in the GH z range but really just because I needed a antenna--not for its direcional properties. Ans if I knew something better for DF work than the loop and smaller, I would buy it if I saw it --like I do no have enough antennas :)
I have seen those videos 1 and 2. As far as I understand, the FHL is the same shielded loop antenna they built in the videos. Only for strong signals I'd need an attenuator. I wanted to buy the antenna you linked, but unfortunately Im not in the US and the closest place I could get it from is GB, which is still too far. I don't want to wait and pay for shipping, so I will try to build it first this weekend and report here how it went...:rolleyes:
 

dlwtrunked

Member
Premium Subscriber
Joined
Dec 19, 2002
Messages
2,436
I have seen those videos 1 and 2. As far as I understand, the FHL is the same shielded loop antenna they built in the videos. Only for strong signals I'd need an attenuator. I wanted to buy the antenna you linked, but unfortunately Im not in the US and the closest place I could get it from is GB, which is still too far. I don't want to wait and pay for shipping, so I will try to build it first this weekend and report here how it went...:rolleyes:
For an attenuator (and that is only if you are real close to the strong signal) , you want an offset one. I carry one with me but have never neededit.
 
Top