Crowdsourcing security for wireless air traffic communications

Protecting the security of the cyber-physical systems that make up the world's critical infrastructures has been a recent hotly debated topic. Legacy wireless communication infrastructure is often an impediment to quickly improving these crucial systems, as cryptographic solutions prove impossible to deploy. In this article, we propose the establishment of a separate verification layer for sensitive wireless data powered by crowdsourced sensors connected to the Internet and apply it to the aviation domain. We first validate the need for independent data verification in air traffic control networks, where all wireless communication is conducted in the clear and thus subject to manipulation. To counter this threat, we develop a comprehensive model for the verification of wireless communication based on massively distributed data collection and outline how it can be used to immediately improve the security of unprotected air traffic control networks. By combining several different methods based on the content and the physical characteristics of aircraft signals, our system is able to detect typical injection, modification and jamming attacks. We further develop a trust model to defend against potential insider threats based on compromised sensors. We illustrate our approach using the crowdsourced sensor network OpenSky, which captures large parts of civil air traffic communication around the globe. We analyse the security of our approach and show that it can quickly, cheaply, and effectively defend against even sophisticated attacks.