A holistic approach for measuring the survivability of SCADA systems

Queiroz Batista da Silva, C 2012, A holistic approach for measuring the survivability of SCADA systems, Doctor of Philosophy (PhD), Computer Science and Information Technology, RMIT University.

Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
da_Silva.pdf Thesis Click to show the corresponding preview/stream application/pdf;... 9.36MB
Title A holistic approach for measuring the survivability of SCADA systems
Author(s) Queiroz Batista da Silva, C
Year 2012
Abstract Supervisory Control and Data Acquisition (SCADA) systems are responsible for controlling and monitoring Industrial Control Systems (ICS) and Critical Infrastructure Systems (CIS) among others. Such systems are responsible to provide services our society relies on such as gas, electricity, and water distribution. They process our waste; manage our railways and our traffic. Nevertheless to say, they are vital for our society and any disruptions on such systems may produce from financial disasters to ultimately loss of lives.
SCADA systems have evolved over the years, from standalone, proprietary solutions and closed networks into large-scale, highly distributed software systems operating over open networks such as the internet. In addition, the hardware and software utilised by SCADA systems is now, in most cases, based on COTS (Commercial Off-The-Shelf) solutions. As they evolved they became vulnerable to malicious attacks.
Over the last few years there is a push from the computer security industry on adapting their security tools and techniques to address the security issues of SCADA systems. Such move is welcome however is not sufficient, otherwise successful malicious attacks on computer systems would be non-existent.
We strongly believe that rather than trying to stop and detect every attack on SCADA systems it is imperative to focus on providing critical services in the presence of malicious attacks. Such motivation is similar with the concepts of survivability, a discipline integrates areas of computer science such as performance, security, fault-tolerance and reliability.
In this thesis we present a new concept of survivability; Holistic survivability is an analysis framework suitable for a new era of data-driven networked systems. It extends the current view of survivability by incorporating service interdependencies as a key property and aspects of machine learning. The framework uses the formalism of probabilistic graphical models to quantify survivability and introduces new metrics and heuristics to learn and identify essential services automatically.
Current definitions of survivability are often limited since they either apply performance as measurement metric or use security metrics without any survivability context. Holistic survivability addresses such issues by providing a flexible framework where performance and security metrics can be tailored to the context of survivability. In other words, by applying performance and security our work aims to support key survivability properties such as recognition and resistance. The models and metrics here introduced are applied to SCADA systems as such systems insecurity is one of the motivations of this work. We believe that the proposed work goes beyond the current status of survivability models. Holistic survivability is flexible enough to support the addition of other metrics and can be easily used with different models. Because it is based on a well-known formalism its definition and implementation are easy to grasp and to apply. Perhaps more importantly, this proposed work is aimed to a new era where data is being produced and consumed on a large-scale. Holistic survivability aims to be the catalyst to new models based on data that will provide better and more accurate insights on the survivability of systems.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Computer Science and Information Technology
Keyword(s) SCADA
Probabilistic Graphical Models
Version Filter Type
Access Statistics: 613 Abstract Views, 3370 File Downloads  -  Detailed Statistics
Created: Wed, 27 Mar 2013, 14:42:17 EST by Keely Chapman
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us