Aarno Labs LLC — Department of Defense SBIR Phase II: SB153-003

Description

Evaluating security mechanisms is a difficult and manual process. To effectively compare different techniques, security analysts use synthetic benchmarks that are typically comprised of simple test cases or a combination of known vulnerabilities in an attempt to provide coverage for different classes of attacks. As such, security mechanisms adapt their techniques to guarantee good coverage for the synthetic benchmarks but provide little guarantees of their efficacy beyond that. To ensure accurate evaluation and comparison of competing security techniques on real-world programs, we need automated techniques for injecting realistic and verifiable vulnerabilities. In phase I, we developed a new technique and system for automatically generating and injecting realistic vulnerabilities to real-world applications. Our proposed system uses targeted symbolic execution to discover program paths that could be used to generate vulnerabilities (e.g., integer overflows). In phase II, we will improve our prototype with enhanced capabilities and improved scalability, robustness and usability. This effort will enhance the prototype's Technology Readiness Level and enable a transition to the DoD and commercial sector. If successful, our proposed system can transform the evaluation of security systems and enable novel pedagogical tools.Evaluating security mechanisms is a difficult and manual process. To effectively compare different techniques, security analysts use synthetic benchmarks that are typically comprised of simple test cases or a combination of known vulnerabilities in an attempt to provide coverage for different classes of attacks. As such, security mechanisms adapt their techniques to guarantee good coverage for the synthetic benchmarks but provide little guarantees of their efficacy beyond that. To ensure accurate evaluation and comparison of competing security techniques on real-world programs, we need automated techniques for injecting realistic and verifiable vulnerabilities. In phase I, we developed a new technique and system for automatically generating and injecting realistic vulnerabilities to real-world applications. Our proposed system uses targeted symbolic execution to discover program paths that could be used to generate vulnerabilities (e.g., integer overflows). In phase II, we will improve our prototype with enhanced capabilities and improved scalability, robustness and usability. This effort will enhance the prototype's Technology Readiness Level and enable a transition to the DoD and commercial sector. If successful, our proposed system can transform the evaluation of security systems and enable novel pedagogical tools.