Dissertation/Thesis Abstract

Comparing the Effectiveness of Resilience Measures
by Matta, John, M.S., Southern Illinois University at Edwardsville, 2014, 59; 1566215
Abstract (Summary)

Because of the ubiquity of networks, both naturally-occurring and engineered, there is a great deal of interest in studying their properties. One such property of a network is its resilience. While there exists a large amount of research on edge-based resilience of homogeneous networks, there is a comparatively smaller amount of study concerning resilience in the face of vertex-based attacks on heterogeneous degree networks. An attempt was made to fill that void by introducing a new measure of network resilience called vertex attack tolerance (VAT). VAT represents the proportion of a network that stays connected, relative to the size of the subset of removed vertices. VAT is motivated specifically to provide better results than existing measures when considering targeted node attacks on scale-free networks. This thesis is a study of vertex attack tolerance and examines many of its properties, including its bounds, its calculability, and its performance as a ranking tool. The attack set that VAT returns is examined for its ability to detect vulnerabilities, as well as for its applicability to graph-based clustering algorithms.

Indexing (document details)
Advisor: Ercal, Gunes
Commitee: Bouvier, Dennis, Yu, Xudong
School: Southern Illinois University at Edwardsville
Department: Computer Science
School Location: United States -- Illinois
Source: MAI 53/06M(E), Masters Abstracts International
Source Type: DISSERTATION
Subjects: Computer science
Keywords: Graph theory, VAT, Vertex attack tolerance
Publication Number: 1566215
ISBN: 978-1-321-23434-3
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest