Explore a 21-minute conference talk from USENIX Enigma 2017 that delves into the vulnerabilities of sensors in safety-critical systems. Learn how Professor Yongdae Kim from KAIST examines the potential for sensor spoofing and manipulation in devices such as self-driving cars, drones, and medical equipment. Discover various attack vectors, including heart rate sensor spoofing, sensor saturation in medical infusion pumps, and gyroscope manipulation in drones. Gain insights into the experimental results of testing MEMS gyroscopes and the implications for drone control. Understand the risks associated with signal injection using EMI and its application to cardiac implantable devices. Conclude with potential defense strategies and open problems in sensor security, providing valuable knowledge for professionals and enthusiasts in the field of cybersecurity and embedded systems.
Overview
Syllabus
Intro
Sensing & Actuation Actuation and decision-making based on sensor dala
Sensor & Security Many prevention and detection mechanisms For malicious network trattos
Attack Vectors of Sensors
Heart Rate Sensor Spoofing
Target: Medical Infusion Pump
Sensor Saturation o New type of sensor spoofing attack using saturation
Infusion Pump and Drop Sensor
Over and Under Infusion
Experimental Setting
Demo (Over-infusion)
Defense for Sensor Saturation
Drone Controlling
Gyroscope on Drone
Experimental Results Found the resonant frequencies of 7 MEMS gyroscopes Not found for 8 MEMS gyroscopes
Shutting Down Drones
Attack Demo
Remote Drone Attack
Signal Injection using EMI
Cardiac Implantable Devices
Application to medical devices
Standard Lead Design
Example
Conclusion
Taught by
USENIX Enigma Conference
