DescriptionIn the current cyberwarfare climate, industrial control systems (ICS) are increas-
ingly becoming focal points of security research as they interconnect, monitor, and
control safety-critical processes. ICS comprise a class of cyber-physical systems (CPS)
across a wide range of domains, including but not limited to the electric power grid,
factory automation, biomedical applications, as well as nuclear reactors. As the inter-
connectivity and accessibility of ICS system components expands, the attack surface for
such systems expands as well. Because these ICS control safety-critical physical pro-
cesses, there is a need for security solutions that have the physical dynamics integrated
into the design process in order to ensure safe operation.
This thesis investigates the security and verification of ICS at different levels of
abstraction. The goal is to bridge the gap between practical security analyses and
sound theoretical approaches to verifying cyber-physical systems. In particular, we
propose to not only leverage the physical properties of an ICS for security purposes,
but to also provide fine-grained hybrid systems modelling of embedded cyber-physical
systems such as a programmable logic controllers (PLCs).
First, this thesis introduces novel and practical security and verification solutions for
ICS that leverage the cyber-physical interdependences between the cyber components
and the underlying physical system. This thesis then explores the feasibility of utilizing
formal methods in the context of complex ICS control processes. Finally, this thesis
introduces a balanced approach to cyber-physical intrusion detection that enforces con-
trol behavioral integrity of a distributed ICS by integrating physical state-estimation
into control-flow monitoring of the associated software.