Supervisor

Suitable for

Abstract

Co-Supervised by Systems Security Lab

Physical-layer fingerprinting of radio signals has historically been used to identify radio devices in environments where the addition of cryptographic signatures may be infeasible, such as in low-power IoT devices or avionic systems with long software life cycles. Satellite communications are a good candidate for this type of identification due to their often limited onboard processing power. Furthermore, certain forms of satellite communications such as weather imagery are designed to be able to be received by anyone, making encryption redundant.

Recent work by the Systems Security Lab has developed an end-to-end pipeline for fingerprinting aircraft using physical-layer characteristics of signals sent using the ADS-B protocol. This system automates the detection, capture, and decoding of messages, then compares against existing samples to determine whether the source of the message is legitimate. Applying fingerprinting to existing larger-scale systems is interesting as it can provide protection against signal forgery and replay at the receiver end without redesigning the entire system.

In this project a student would adapt the existing fingerprinting pipeline to work with satellite communications. This will require changes to the message capture components of the pipeline in addition to the fingerprinting model itself. This pipeline can then be assessed for its effectiveness at correctly identifying devices, and identifying signals injected by an attacker.

Useful links: - Existing fingerprinting pipeline: https://github.com/ssloxford/auto-phy-fingerprint - Paper introducing physical-layer radio fingerprinting: https://ethz.ch/content/dam/ethz/special-interest/infk/inst-infsec/system-security-group-dam/research/identification/CSUR-danev.pdf - TDOA satellite fingerprinting: https://dl.acm.org/doi/pdf/10.1145/3448300.3469132