Optimising the OpenSky network

Working paper

In recent years the availability of cheap and readily available radio transceivers has brought the lack of security in modern avionic communication systems into the spotlight. The easiest way to detect attacks on these systems is to use centralised sensor networks (such as OpenSky) to detect discrepancies in data collected by nodes. For large systems such as global air traffic the computational demands can be prohibitive, and can prevent the use of more sophisticated analytics. This report presents three methods for optimising these networks: constraint satisfaction, a greedy approach, and a genetic algorithm. While all solutions produced meaningful optimisations, the speed of the greedy approach means that it is the only solution which can be feasibly run without first clustering data sets, making it the most effective in practice.

Authors: Seymour, W

• Publication status: Not published
• Peer review status: Not peer reviewed
• Host title: CDT Technical Paper Series
• Series: CDT Technical Paper
• Publication date: 2018-01-01
• Paper number: 01/18
• Keywords: CDT