Performance Evaluation and Protection Management for Wireless Networks

Show full item record

Title: Performance Evaluation and Protection Management for Wireless Networks
Author: Agarwal, Avesh Kumar
Advisors: Arne Nilsson, Committee Member
Carla D Savage, Committee Member
Wenye Wang, Committee Chair
David Thuente, Committee Member
Mihail Sichitiu, Committee Member
Abstract: Wireless networks are more vulnerable to attacks compared with wired networks due to their open and shared medium. A number of security mechanisms are available for providing different levels of protection, but with a side effect that is performance degradation. Given a wide range of real-time and non real-time applications, there is a critical demand for providing better performance in wireless networks. Motivated by this demand and the limitations of current research, we devote this dissertation to performance evaluation, and protection management in wireless networks. In this dissertation, we first study the performance impact of security mechanisms at different networking layers in wireless LANs, on a testbed with IP mobility support. One of the important observations is that statically configured security mechanisms may not be used to achieve a satisfactory performance. Motivated by the results of our first study, we propose a Self-TunEd Performance and Protection (STEP2) management framework, which is a generalized architecture for balancing the protection strength and performance for wireless clients based on network and user requirements. By implementing STEP2 in a wireless LAN testbed, we demonstrate the benefits of STEP2 such as significant reduction in delay and packet losses in various scenarios. With the growing deployment of multi-hop networks, we next extend our research to wireless mesh networks. Specifically, we investigate the performance impact of attacks in wireless mesh networks (WMNs). We study path-based denial of service (DoS) attacks as they are easy to carry out and can cause significant performance degradation. The measurements obtained in a wireless mesh testbed demonstrate the interaction between a set of factors, such as link qualities, physical diversity, packet size and rate, and the intensity of path-based DoS attacks. We believe that the comprehensive results in this dissertation will shed new lights on the performance analysis of secure wireless networks.
Date: 2009-08-07
Degree: PhD
Discipline: Computer Science

Files in this item

Files Size Format View
etd.pdf 9.854Mb PDF View/Open

This item appears in the following Collection(s)

Show full item record