Citation
Hall, John M.,
An Investigation into Immune-Based Intrusion Detection, Master's Thesis, University of Idaho, 2003.
Abstract
This research evaluates and implements immune-based intrusion detection techniques in a research intrusion detection system. In this process, we introduce a generalized architecture and two prototype implementations. We also introduce the dynamic affinity, greedy detector set selection, and dynamic maturation time algorithms to improve the efficiency and applicability of immune-based techniques. We compare the first prototype to a simple misuse-based intrusion detection system and the second prototype to an advanced probabilistic method. In each case, we find the tested systems to be comparable. We also perform an evaluation of the dynamic affinity and greedy detector set selection algorithms. We find that both improve the quality of detector sets, but that the greedy detector set selection algorithm is inefficient for creating large detector sets. In more general terms, we show that immune techniques can be very successful in intrusion detection systems.
Download
(
Full Text in PDF)
BibTex Entry
@mastersthesis{Hall2003:Thesis,
author = "John M. Hall",
title = "An Investigation into Immune-Based Intrusion Detection",
school = "University of Idaho",
year = "2003",
url = "http://www.johnmhall.net/research/immunology/Thesis/thesis.html" }
