Tagged: drones

Decoding 5GHz NTSC Video from Drones with a HackRF, DragonOS and SigDigger

Over on his YouTube channel Aaron has uploaded a video showing how we can SigDigger to decode analog NTSC video from a drone camera which is transmitted at 5.7 GHz. SigDigger is a rapidly evolving SDR program for Linux and MacOS that has a lot of built in functionality for inspecting signals in more depth. Although not specifically designed for it, the Symbol Stream viewer in SigDigger can be used to display NTSC Analog Video. Aaron writes:

For the most part, the older an analog modulation is, the easier it is to get basic results when decoding. TV receivers were rather dumb back in the day, basically fast fax machines glued to an off-band FM radio receiver. Receiver circuits were also slow, and the signal had lots of invisible blank spaces in the borders so that the cheapest TVs could switch to the next line in time. The invention of Teletext leveraged those blanks in order to carry digital information and color information was embedded as an additional narrowband signal in the gaps in the spectrum.With this in mind I wanted to take a look at decoding analog video transmissions from drones. While some drones have moved to more effective digital compression and channel transmission technologies allowing for high definition video, there’s still drones using RC-like communications and the FPV video link is pure FM-modulated NTSC.

Searching the internet provided few results on how I could go about using low cost equipment, such as the HackRF One, to decode drone feeds. After an extensive search I decided to start looking at Linux based software defined radio applications I was already familiar with. By chance I happened to be working with SigDigger, a free digital signal analyzer. It has been discussed on RTL-SDR.com and more recently on Signal Lounge (https://signal-lounge.com/2020/05/05/sigdigger-for-signal-analysis/). It is also included in my own creation, DragonOS (https://sourceforge.net/projects/dragonos-lts/)

After a brief email exchange with the developer it was brought to my attention that visualizing analog video transmission is possible in SigDigger (although with no color information, of course). Since SigDigger supports the HackRF and the HackRF provides coverage in the 5ghz band, it was now possible for me to try to decode a 5ghz drone video feed. I’ve documented the process and my results on my YouTube channel. I should point out that this is currently a side feature of SigDigger and currently lacks synchronization. The symbol view area I used in the video is not made for this. It is meant to display symbols and symbols patterns which, due to its behavior, can incidentally show the contents of analog TV and weather faxes with lots of manual adjustments.

While the SigDigger developer makes mention of plans to include an embedded generic analog TV viewer and possibly add the ability to automatically sync video, there’s currently no timeframe on when that might become available.

SigDigger Decoding NTSC Video from a Drone Camera
SigDigger Decoding NTSC Video from a Drone Camera

DragonOS LTS SigDigger demodulating a 5 GHz analog video/FPV drone link (HackRF One, SigDigger)

We note that if you're interested in PAL/NTSC decoding, there is also the excellent TVSharp plugin for SDR# available.

Tracking Tagged Orangutans in the Bornean Jungle with Drones, GNU Radio and an Airspy Mini

Due to various human activities causing the environmental destruction of it's habitat, the Orangutan is now classed as a critically endangered species. In addition to being endangered, Orangutans face another problem in that they are often captured and sold as pets due to their intelligence and cuteness.

To combat these problems, NGOs, charities and rescue centers have been using RF tags on rehabilitated Orangutans that have released back into the wild. The RF tag regularly transmits a data-less pulse at VHF frequencies which is then typically tracked using direction finding equipment such as a directional Yagi antenna. The range is only approximately 200-400m. 

In order to try and alleviate the range issue Dirk Gorissen has been working on creating a drone based system that could detect the VHF transmission and create a heatmap of Orangutan positions. The first iteration of his system uses an RTL-SDR, Odroid and lightweight loop antenna. A simple Python script then monitors the spectrum and logs the drones current location, altitude, speed and heading when a pulse is detected. Tests confirmed that the signal was able to be detected from the sky, but unfortunately the drone was eventually crashed and lost before it could be properly used.

In his second try a few years later, Dirk used a larger drone and switched SDRs to an Airspy Mini with preamp. The pulse detection code was also improved by using GNU Radio to create a DSP algorithm combining peak detection, cross correlation with a known template of the signal, and a phase locked loop. Visualization and data transfer is achieved through react.js and a Flask web server running on the drones WiFi hotspot. This time with the new drone and system Dirk was able to successfully detect and locate several Orangutan's on various flights, despite noting that some RF tags appeared to be glitchy.

Orangutan Detected with Drone, Airspy Mini and GNU Radio.
Orangutan Detected with Drone, Airspy Mini and GNU Radio.
Drone used in the experiment
Drone used in the experiment