Using Process Invariants to Detect Cyber Attacks on a Water Treatment System
Abstract
An experimental investigation was undertaken to assess the effectiveness of process invariants in detecting cyber-attacks on an Industrial Control System (ICS). An invariant was derived from one selected sub-process and coded into the corresponding controller. Experiments were performed each with an attack selected from a set of three stealthy attack types and launched in different states of the system to cause tank overflow and degrade system productivity. The impact of power failure, possibly due to an attack on the power source, was also studied. The effectiveness of the detection method was investigated against several design parameters. Despite the apparent simplicity of the experiment, results point to challenges in implementing invariant-based attack detection in an operational Industrial Control System.
Origin | Files produced by the author(s) |
---|
Loading...