Frederick Rawlins, Richard Baker and Ivan Martinovic (University of Oxford) Presenter: Frederick Rawlins
Satellites in Geostationary Orbit (GEO) provide a number of commercial, government, and military services around the world, offering everything from surveillance and monitoring to video calls and internet access. However a dramatic lowering of the cost-per-kilogram to space has led to a recent explosion in real and planned constellations in Low Earth Orbit (LEO) of smaller satellites.
These constellations are managed remotely and it is important to consider a scenario in which an attacker gains control over the constituent satellites. In this paper we aim to understand what damage this attacker could cause, using the satellites to generate interference.
To ground our analysis, we simulate a number of existing and planned LEO constellations against an example GEO constellation, and evaluate the relative effectiveness of each. Our model shows that with conservative power estimates, both current and planned constellations could disrupt GEO satellite services at every groundstation considered, albeit with effectiveness varying considerably between locations.
We analyse different patterns of interference, how they reflect the structures of the constellations creating them, and how effective they might be against a number of legitimate services. We find that real-time usage (e.g. calls, streaming) would be most affected, with 3 constellation designs able to generate thousands of outages of 30 seconds or longer over the course of the day across all groundstations.