Explore a novel detection system for low-level electromagnetic interference attacks on sensor systems in this IEEE conference talk. Learn about the vulnerabilities of sensor systems used in various applications, from home automation to critical infrastructure. Discover how attackers can manipulate sensor outputs through intentional electromagnetic interference and the potential consequences of such attacks. Delve into the proposed detection method, which utilizes sensor modulation to identify adversarial signals. Understand the working mechanism, detection guarantees, and implementation of this approach in microphone and temperature sensor systems. Gain insights into the effectiveness and robustness of this detection mechanism, which offers a cost-effective and easy-to-implement solution for enhancing sensor system security.
Detection of Electromagnetic Interference Attacks on Sensor Systems
Overview
Syllabus
Intro
Sensor System: Vulnerabilities
Our Approach: Turn the Sensor ON
Our Approach: Physical Quantity is Constant
Our Approach: Random Bias Voltage
Our Approach: Two Requirements
Our Approach: When An Attack Happens
A Special Case: Non-constant Physical Quantity
Implementation: Temperature Sensor System
Implementation: Deploy Our Approach
Conclusion
Taught by
IEEE Symposium on Security and Privacy
Tags
Reviews
5.0 rating, based on 1 Class Central review
-
The class on "Detection of Electromagnetic Interference (EMI) Attacks on Sensor Systems" was great and highly educational. It explored the methods used to detect and mitigate EMI attacks, which can disrupt or manipulate sensor systems critical to various industries, including automotive and healthcare. We learned about different techniques to identify interference, such as monitoring sensor data patterns and implementing shielded hardware designs. The class also covered countermeasures to protect sensors from EMI, such as filtering noise and isolating signals, providing a strong foundation for securing sensor systems against potential electromagnetic threats.
