Protecting the Satellite Ecosystem

Abstract

Satellites are becoming increasingly intertwined in our daily lives, whether we are aware of it or not. As the importance and prevalence of satellites continues to increase, so do the implications on the privacy, security, and safety of humans on Earth and in space. These risks are not limited to a particular type of satellite nor a particular adversary. They span the entire satellite ecosystem. For example, state adversaries may seek to damage satellites launched by a competing country for financial gain, or to destroy a portion of satellite-facilitated critical infrastructure as part of cyber warfare. Or, even without a security compromise, individuals could purchase increasingly affordable commercial satellite images for interpersonal surveillance. This dissertation identifies and addresses security and privacy challenges across the satellite ecosystem that have previously received little attention. First, this work surveys the possible threats of commercial imaging satellites to individual privacy. Commercial imaging satellites pose both possible benefits and privacy harms, and this work explores these tradeoffs through a survey of 99 participants, followed by a discussion of possible solutions that still maintain utility benefits. Second, this work looks at the importance of smaller satellites and argues why they must still practice good security. Because of the cyberphysical effects of satellites, one insecure satellite has the capacity to damage other satellites in the ecosystem. This interconnectedness necessitates careful analysis of security and other possible harms among all satellite operators of every satellite in orbit. This work puts forth possible threats posed by university smallsats, along with barriers to building them securely through a combination of in-depth interviews and open-source code analysis. Lastly, this work focuses on emerging threats to more complex satellites or fleets of satellites. This work identifies the risks of a compromised peripheral on a large complex satellite. This work demonstrates through simulation that current resiliency techniques do not handle the case when a satellite peripheral is malicious rather than just faulty. Then, the final portion of this work proposes and experimentally tests a novel technique that identifies and takes action on malicious peripherals. Together, this work makes steps toward a safer satellite ecosystem that benefits those on the ground.