Files in this item



application/pdfHaohui_Mai.pdf (2MB)
(no description provided)PDF


Title:Building abstractions for fast, secure, reliable computer systems
Author(s):Mai, Haohui
Director of Research:King, Samuel T.
Doctoral Committee Chair(s):King, Samuel T.
Doctoral Committee Member(s):Godfrey, Philip B.; Caesar, Matthew C.; Zeldovich, Nickolai
Department / Program:Computer Science
Discipline:Computer Science
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):computer system
data plane analysis
network troubleshooting
Boolean satisfiability
parallel Web browser
Web page decomposition
multi core
mobile security
programming by contracts
automatic theorem proving
Abstract:Modern computer systems play important roles in our society and everyday lives. Their performance, security and reliability are of critical importance. Real-world computer systems, however, occasionally suffer from performance degradation, security exploits, and poor reliability, because of the lack of efficient automatic analyses. This dissertation introduces a new methodology for building efficient automatic analyses for real-world computer systems through identifying and designing proper abstractions. It demonstrates the methodology within the context of three real-world computer systems: detecting net- work defects at the data plane level, exploiting data parallelism in web pages, and formally verifying security invariants in operating system kernels. This dissertation presents the design, implementation, and evaluation of the above systems, and shows that choosing the proper set of abstractions is an essential step to constructing efficient automatic analyses for real-world computer systems. Moreover, these analyses can become valuable tools to improve the performance, security and reliability of computer systems.
Issue Date:2013-08-22
Rights Information:Copyright 2013 Haohui Mai
Date Available in IDEALS:2013-08-22
Date Deposited:2013-08

This item appears in the following Collection(s)

Item Statistics