Accelerometers and gyroscopes are commonly used sensors which measure varying forms of motion. Accelerometers measure a type of acceleration typically known as “g-force", while gyroscopes measure angular momentum. In order to make these sensors usable in embedded systems, compact designs for these sensors have been developed in the form of MicroElectroMechanical Systems (MEMS). MEMS devices are packaged in integrated circuit (IC) packages and, as their name suggests, contain very small mechanical and electrical components. MEMS and piezoelectric sensors can be found in many of the systems and devices mentioned above, some of which rely one of these sensors to make mission critical decisions. Recent work has shown how gyroscope sensors can be maliciously stimulated to produce output using acoustic pressure waves. Other work has shown how this acoustic stimulation can cause systems which rely on their gyroscopes to behave erratically and sometimes fail.
Smart sampling strategies to avoid malicious interference
The inventors present a technology for preventing malicious sounds from detonating airbags, crashing drones, and other cyberphysical and Internet of Things (IoT) systems. All known acoustic attacks involve generation of sounds as the resonant frequency of the physical material found inside MEMS devices, so the inventors developed the defense mechanisms using novel sampling strategies to ensure that the interference does not cause constant or semi-constant effects on the output of acceleration sensors
- IoT systems
- The only solution except physical damping of sound
- Timothy Trippel
- Ofir Weisse
- Peter Honeyman