Spoofing Attack on Ultrasonic Distance Sensors Using a Continuous Signal

Area of Science:

• Robotics and Autonomous Systems
• Cybersecurity
• Sensor Technology

Background:

• Ultrasonic distance sensors are crucial for autonomous vehicles, enabling object avoidance and altitude measurement.
• The reliability of these sensors is vital for vehicle operation, making them targets for adversarial attacks.
• Previous attacks, like denial-of-service and limited spoofing, required specific sensor knowledge.

Purpose of the Study:

• To present a novel spoofing attack on ultrasonic distance sensors.
• To demonstrate an attack that does not require prior knowledge of sensor operation or timing.
• To evaluate the attack's feasibility and impact in both lab and real-world scenarios.

Main Methods:

• Exploited a vulnerability in the ultrasonic sensor's receiver.
• Utilized a constant noise input to generate a fake ultrasonic pulse.
• Validated the attack in a controlled lab testbed and a drone-based real-world setup.

Main Results:

• Successfully demonstrated a spoofing attack capable of altering ultrasonic sensor measurements.
• The attack requires only a constant signal, eliminating the need for sensor-specific timing or location data.
• Experimental results confirmed interference with autonomous vehicle operation.

Conclusions:

• The novel attack poses a significant cybersecurity risk to autonomous vehicles.
• The technique can be adapted as a defensive measure to control unauthorized vehicle movement.
• Further research into ultrasonic sensor security is warranted.