Leveraging Drones' Propeller Vibration as a Covert Channel for Authentication

We are currently witnessing a transformative era in which drones are becoming deeply embedded across a diverse range of sectors and applications. These unmanned aerial vehicles offer significant advantages, improving safety in high-risk environments such as military operations and increasing efficiency in tasks such as logistics and transportation. Drones were originally developed for military purposes with the primary objective of gaining strategic and tactical advantages. Over time, they have evolved and found utility in various fields, making work easier, more precise, and safer. However, the growing ubiquity of drones has also attracted the attention of malicious actors. These adversaries seek to exploit drones by compromising their systems, impersonating legitimate devices, seizing control, or infiltrating sensitive areas. Their proximity to critical infrastructure poses serious security risks, which prompts researchers to seek effective countermeasures. To date, most proposed solutions have been centered on cryptographic challenge-response authentication mechanisms. These mechanisms typically rely on lightweight cryptographic algorithms to accommodate the limited computational resources of drones. However, as attackers gain access to increasingly powerful computing capabilities, these lightweight algorithms become more vulnerable. For example, symmetric cryptographic keys can now be more easily obtained or broken. In response to this growing concern, this thesis explores an alternative approach that goes beyond traditional cryptographic methods. Specifically, we investigate the use of the physical characteristics of drones as a means of authentication. The proposed solution aims to mitigate the vulnerabilities associated with standard cryptographic tools by leveraging each drone’s unique physical signature. Our research focuses on exploiting propeller vibrations as a covert channel for drone authentication. Using radar systems, we capture various physical parameters, such as the propeller frequency, rotational speed, and vibration patterns. Because these characteristics are inherently tied to the physical structure and design of the drone, they serve as distinctive signatures that are difficult to replicate. By analyzing these physical traits, we propose a novel authentication protocol that uses these measurable features to verify the identity of a drone. This method not only enhances security but also provides a complementary layer of defense alongside existing cryptographic techniques.