Abstract
Many-core systems are susceptible to attacks launched by thermal covert channel (TCC) attacks. Detection of TCC attacks often relies on the use of threshold-based approaches or variants, and a countermeasure to thwart the channel can be applied only after an attack is deemed to be present. In this article, we describe a direct sequence spread spectrum (DSSS)-based TCC, where its thermal data are modulated by a pseudo-random bit sequence. Unfortunately, such DSSS-based TCC has an extremely low signal strength that the signal is nearly indistinguishable from the noise and thus cannot be detected by any existing threshold-based detection methods. To combat this stealthy TCC, we propose a novel detection scheme that lets the received signal pass through a differential filter where irrelevant frequency components occupied mainly by the noise gets eliminated and the filtered signal is next compared against a threshold for successful detection. Experimental results show that the DSSS-based TCC can effectively survive detection by the existing detection methods with its BER as low as 4%. In contrast, with the proposed detection and countermeasure applied, the detection accuracy jumps to 89%, and the BER of the DSSS-based TCC soars to 50%, which indicates that the TCC is practically shut down.
Funding Information
  • National Natural Science Foundation of China (61971200)
  • Zhejiang Lab (2021LE0AB01, 2021PC0AC01)
  • Major Scientific Research Project of Zhejiang Lab (2021LE0AC01)
  • Open Research Grant of State Key Laboratory of Computer Architecture Institute of Computing Technology, Chinese Academy of Sciences (CARCH201916)
  • Key Technologies Research and Development Program of Jiangsu (BE2021003)
  • Key Laboratory of Big Data and Intelligent Robot (South China University of Technology), Ministry of Education
  • National Key Research and Development Program of China (2019QY0705)

This publication has 35 references indexed in Scilit: