Controllability and Observability of Risk and Resilience in Cyber-Physical Cloud Systems

Abstract

Effective management of risk and resilience in a dynamic cyber-physical system is essential for ensuring successful completion of missions by minimizing the adverse impact of attacks and physical failures. With the accurate risk assessment and efficient resilience control of security events and operations, a cyber-physical system can keep performing satisfactorily by adapting to the dynamic changes occurring due to various cybersecurity events and operations, such as exploiting vulnerabilities, detecting intrusions, and recovering compromised nodes. To the best of our knowledge, this book chapter is the first one to present a model with system state equations of linear and non-linear, based on cybersecurity parameters such as cyber assets’ vulnerabilities, criticalities, dependencies, influences, attack types, intrusions, recovery rate, patching rate, normal and compromised nodes. Using this model, this book chapter describes how to apply the controllability and observability aspects of linear/non-linear systems to manage cybersecurity risk and resilience of cyber-physical systems. The purpose of employing controllability is to steer a system from an abnormal security state to a normal security state. That is, by implementing recovery and resilience operations on compromised nodes and assets of a system, it is steered from an abnormal state with compromised nodes towards a state with a fewer or no compromised nodes. Observability is used to determine the system security state by having appropriate cyber output measurements. The challenges for implementing controllability and observability are discussed. An example is provided to illustrate how controllability could be used to achieve resilience within a network.