Jun Sun, Singapore University of Technology and Design
Software Engineering Techniques for Cyber-Physical Systems


Cyber-Physical Systems (CPS) are often safety-critical yet they are challenging to analyze. In this talk, I will introduce our recent attempts on solving research problems associated with safety and security of CPS through techniques developed in software engineering community. Firstly, I will show a practical approach which applies program mutation and machine learning to attest CPS, so as to detect code-modification attacks. Secondly, I will present a theoretical analysis on existing concolic testing strategies, which leads to a practical "optimal" strategy. We show that the new strategy significantly outperforms existing strategies on testing software programs including control programs in CPS. The tutorial is based on our papers accepted at IEEE S&P 2018 and ICSE 2018.