Cross-Layer Secure and Resilient Control of Delay-Sensitive Networked Robot Operating Systems

Zhiheng Xu, Quanyan Zhu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

A Robot Operating System (ROS) plays a significant role in organizing industrial robots for manufacturing. With an increasing number of the robots, the operators integrate a ROS with networked communication to share the data. This cyber-physical nature exposes the ROS to cyber attacks. To this end, this paper proposes a cross-layer approach to achieve secure and resilient control of a ROS. In the physical layer, due to the delay caused by the security mechanism, we design a time-delay controller for the ROS agent. In the cyber layer, we define cyber states and use Markov Decision Process to evaluate the tradeoffs between physical and security performance. Due to the uncertainty of the cyber state, we extend the MDP to a Partially Observed Markov Decision Process (POMDP). We propose a threshold solution based on our theoretical results. Finally, we present numerical examples to evaluate the performance of the secure and resilient mechanism.

Original languageEnglish (US)
Title of host publication2018 IEEE Conference on Control Technology and Applications, CCTA 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1712-1717
Number of pages6
ISBN (Electronic)9781538676981
DOIs
StatePublished - Oct 26 2018
Event2nd IEEE Conference on Control Technology and Applications, CCTA 2018 - Copenhagen, Denmark
Duration: Aug 21 2018Aug 24 2018

Publication series

Name2018 IEEE Conference on Control Technology and Applications, CCTA 2018

Other

Other2nd IEEE Conference on Control Technology and Applications, CCTA 2018
Country/TerritoryDenmark
CityCopenhagen
Period8/21/188/24/18

ASJC Scopus subject areas

  • Aerospace Engineering
  • Control and Optimization
  • Automotive Engineering
  • Safety, Risk, Reliability and Quality

Fingerprint

Dive into the research topics of 'Cross-Layer Secure and Resilient Control of Delay-Sensitive Networked Robot Operating Systems'. Together they form a unique fingerprint.

Cite this