[Colloq] MS in IA Thesis Presentation - Guided Vulnerability Analysis of Arbitrary Binary Objects with Dynamic Invariant Detection - Ryan Rickert - 4/26, 3pm, 366 WVH

[Jessica Biron]

Guided Vulnerability Analysis of Arbitrary Binary Objects with Dynamic Invariant Detection 

Ryan Rickert 

Friday, April 26th 
3:00pm 366 WVH 

Abstract: 

The goal of this thesis project is to develop a unique and systematic approach to vulnerability analysis 
via dynamic invariant detection. The targets for analysis are executable programs for which source code 
and debugging information are not available. A variety of techniques exist for both vulnerability analysis 
and binary analysis. Unfortunately, there are few solutions that address both. Most vulnerability analysis 
tools and techniques require access to source code, or at least code that was compiled with debugging 
symbols. On the other hand, most binary analysis tools and techniques are either not automated, not 
tailored to vulnerability analysis, or too narrowly focused on specific classes of vulnerabilities. 
We present a procedural approach for leveraging the Daikon Dynamic Invariant Detector to extract invariant 
properties of programs using only information dynamically gathered from binary executions.