Visible to the public Side-Channel Attack Resistance—2022Q3Conflict Detection Enabled

Submitted by heechul on Sun, 10/09/2022 - 2:26pm

PI(s): Heechul Yun

HARD PROBLEM(S) ADDRESSED: Resilient Architectures

This project is developing OS and architecture techniques to defend against potential microarchitectural side-channel attacks on embedded computing platforms for safety-critical systems. The project covers the hardware problems of (1) resilient architectures (primary) and (2) security-metrics-driven evaluation, design, development and deployment.

PUBLIC ACCOMPLISHMENT HIGHLIGHTS:

In this quarter, we have made the following progress.

  • We have continued to develop a new microarchitectural attack and an OS kernel-level defense technique.
  • Our paper, "DeepPicarMicro: Applying TinyML to Autonomous Cyber Physical Systems," was published at RTCSA 2022. It describes our optimization of a CNN based end-to-end deep learning model for an autonomous RC car on a micro-controller. The preprint of the paper got a media coverage (hackers.io).

PUBLICATIONS FROM THE QUARTER

  • [C] Michael Bechtel, Qitao Weng and Heechul Yun. DeepPicarMicro: Applying TinyML to Autonomous Cyber Physical Systems. IEEE International Conference on Embedded and Real-Time Computing Systems and Applications (RTCSA), August 2022 [paper] [slides] [code] [hackers.io]

COMMUNITY ENGAGEMENTS:

  • Dr. Yun is serving as the general chair of RTAS'23.

EDUCATIONAL ADVANCES:

  • Dr. Yun ran a summer camp program on autonomous RC cars (based on our DeepPicar platform) for rural high-school students. The feedback from the students was very positive.
Groups: