Protecting Automotive Analog Sensor Security
Associate Professor of Electrical Engineering and Computer Science, College of Engineering
Research Investigator, Electrical Engineering and Computer Science and Intermittent Lecturer in Electrical Engineering and Computer Science, College of Engineering
To make driving decisions, connected and automated vehicles (CAVs) are critically dependent on the trustworthiness of sensors. While digital security largely depends on deployment of existing science and engineering techniques, analog security remains an open research problem with few tangible advances. Recently discovered vulnerabilities show the ease with which an adversary can disrupt the availability and integrity of sensor data via radio frequency and acoustic interference. The proposed research will investigate how to protect automotive sensors at the analog layer, as well as design patterns, to enable systems to better assess the trustworthiness of input from untrusted sensors.
This proposed effort is singular in its intention to protect the sensors themselves, rather than simply the firmware or software. A key outcome will be the innovation of automotive systems that can remain trustworthy even if the sensors remain inexpensive and insecure. This is in contrast to the conventional approach: attempting to build crash-proof components. Since all components can be hacked, our more end-to-end approach focuses on how to build safe and reliable CAVs despite the presence of hackable components.