Abstract
Critical Information Infrastructure is an important part of any country’s functioning. It includes
most if not all of the country’s important systems and subsystems that allow the
country to operate effectively and provide critical services to its citizens. These services
comprise important infrastructure such as telecommunications and energy, without which
the country would grind to a halt and fall into chaos. The advent and widespread use of the
Internet has created a world in which information is made easily accessible worldwide.
Much of this information is private however, and needs to be protected from exploitation.
Criminals have devised means to circumvent the privacy of information so as to interfere
with the information, with the intention of possibly damaging or making this information
unavailable. This poses a great threat with regard to the security of Critical Infrastructures,
as attacks on these infrastructures may cause significant damage to them. Malware and
other threats can be used by criminals to compromise a targeted system in various ways.
Critical Information Infrastructure Protection is aimed at the protection of a country’s valuable
Critical Infrastructure and must be focused on a comprehensive approach that aims to
protect CIs as a single system made up of various smaller subsystems. A possible approach
for achieving this protection is the creation of an artificial immune system capable of acting
like the biological immune system in terms of detecting potential threats. This AIS would
need to make use of various immunological models to detect and deal with unknown
threats, and be made up of learning agents capable of adaption and learning, thus allowing
the system to react to new threats and remember old ones. The AIS would have to be implemented
in a manner allowing it to operate independently of the operating system so that
it is itself protected from attacks and infection.
In this dissertation, the VAIS model is introduced for the creation of an AIS capable of protecting
a computer system from both known and unknown threats. The VAIS model is designed
to be deployed across a massively distributed computer system for the protection of
critical information infrastructures. The VAIS model is a layered model broken up into three
main layers: the Immune Process layer, the Immune Agent layer, and the Distributed Agent
Management Network. These three layers are designed to work together to define the basis
of creating a distributed AIS, covering all aspects from the definition of the immune processes
and functions that will be simulated, to the creation of specifically tasked agents and
the management thereof.
M.Sc. (Computer Science)