Analysis of the Potential Malicious Code on Control Systems

Abstract

After the 911 terrorism attacks, the America government thoroughly investigated the infrastructure environment and found that the automated control system is a vulnerable point. In order to ensure the control system programming security, the malicious program issue must be investigated. Thus, our research focuses on the control system malicious code. This paper classified the malicious code based on control system characteristics. We identified the potential malicious code classifications, such as time-dependent, data-dependent, behavior-dependent, input-dependent, violate a certain theorem, deviation’s error, isolated codes, change of privilege/authentication, anomalous file access, access of text segment as data, and so on. Finally, we develop an example to construct a malicious program on automated control system according to our classification. It demonstrates that the malicious code written by an insider and planted into the control system program is possible. So, this research is the first step towards detecting the malicious code on the control systems.