Date of Completion

3-29-2013

Embargo Period

3-28-2013

Keywords

fault performance monitoring wlan

Major Advisor

Bing Wang

Associate Advisor

Reda A. Ammar

Associate Advisor

Lester Lipsky

Associate Advisor

Jun-Hong Cui

Associate Advisor

Zhijie Shi

Field of Study

Computer Science and Engineering

Degree

Doctor of Philosophy

Open Access

Open Access

Abstract

Air sniffering is a widely-used and effective technique to monitor access points in WLANs. This technique, however, requires a large number of sniffers and generates a large amount of data. These challenges can be overcome by channel sampling, where each sniffer samples the network traffic by visiting multiple channels periodically. In the first part of the dissertation, we address an important problem in channel sampling, namely, how to select channels for sniffers to reduce monitoring cost. Specifically, we study two channel selection problems. Both of them require that each AP be monitored by at least one sniffer, and in addition, one problem minimizes the maximum number of channels that a sniffer listens to, while the other minimizes the total number of channels that the sniffers listen to. We propose three algorithms, one based on integer program, LP-relaxation, and greedy heuristic, to solve each problem. The performance of the algorithms is evaluated extensively using real-world traces.

Download
COinS