So it’s been a looong time since my first post and while I’ve done occasional bits of work on projects I have largely been a bit too busy to do anything lately. This is, other than life just getting in the way and time going way too fast, because in July last year I started as a PhD student and it’s keeping me very busy!
While this situation isn’t going to be changing in the foreseeable future, I have more recently been digging back into my projects. I think the best way to try to continue making progress is to keep up momentum, and a great way to do that is to do a bit of writing about what I’m doing on the side. My posts won’t be particularly polished but they will be more frequent – I’ll be aiming for about a post every 2 weeks, and hopefully in doing so, I make enough progress to have things to talk about. Hopefully posting also helps me order my thoughts better if and when I encounter some difficulties in my projects, and one day, hopefully, I can get some good feedback and input from others too!
So in the absence of any particular personal project updates at this stage, and to commemorate my 1 year anniversary, I want to talk about my PhD project.
Use of Radio Astronomy for Satellite Detection
The overarching theme of my PhD is to use radio astronomy techniques to detect and characterise (human-made) satellites in orbit – mainly Low Earth Orbit (LEO) which is generally taken to be up to ~2,000km altitude. I’m using a small purpose built radio interferometer to capture data, which I can then process in a variety of ways to tease out the interesting information.
Radio interferometry involves the use of an array or network of separate antennas or dishes which capture data simultaneously, and have their feeds cross-correlated against those of all the other feeds, often in real time as the data is being captured. These cross-correlation products (carried out in the frequency domain) are generally referred to by the name visibilities, and have the seemingly magic ability – at least to an absolute beginner when I was first hearing about them in July last year – to be turned into an image of the sky. The angular resolution of the resulting image is based on the separation between the antennas, rather than based only on the size of the individual elements in the array (although this is of course also a factor in the resulting data and image).
Satellites in LEO can be picked up in the radio spectrum in a couple of different ways:
- Detecting transmission directly emitted by the satellite – which can be further characterised as intentional, i.e. telemetry, and unintentional, i.e. poor design or faulty components emitting RFI (radio frequency interference). This is referred to as Passive RF
- Detecting reflected transmission from terrestrial transmitters, such as commercial FM radio which lights up the satellite and allows reflection of that signal to be detected. This is referred to as Non-coherent Passive Radar, and works in essentially the same way as Passive RF, but with a much lower in flux
- As above, but cross-correlating the received sky signal against a reference signal from the illuminator (such as the FM radio station) which hugely improves sensitivity, but is processed in a very different way and results in different outputs – range and radial velocity of the object, rather than angular position as with the other methods. This method is referred to as Coherent Passive Radar
So far, I’ve had some amount of success with the first two, making confident detections of satellites using each, and have made some progress towards the third but no actual result or detection yet. There’s a lot of groundwork to be done before making a detection, and a large portion of the work I’m involved in is commissioning activity for the instrument I’m using since it was newly constructed (and I helped deploy the majority of the hardware late last year). Any new system requires a fair bit of work to get from deployed to properly up and running and of late there’s been a few hardware issues needing debugging and replacement of some components, holding up data collection and further development. Calibration can also be deceptively difficult with an interferometer, but that’s a more detailed discussion for another day.
All of the hardware issues seem to be sorted as of last week, so hopefully I can start making more development and progress again in the near future, and will provide some extra updates in due course.