Automated Repair of Buffer Overflow

Student: Christian Ellwood

Major: Computer Science

Mentor: Dr. Kris Ghosh 

Department: Computer Science 

Automated Repair of Buffer Overflow

The project is to define a process in which software that has a potential vector of attack of memory manipulation, will be able to detect the vulnerability and take steps to prevent an attacker to gain access to the system. This process of detection of a software vulnerability will be represented using language that is represented by mathematical expression (consisting of Boolean logic). Using these mathematical expressions, the process will design constraints by expressing the specifications that the software must abide by in order to continue operation. Another major goal of the project is to ensure that any software that uses the process incurs minimal memory usage and that the computational power required to run is not a significant strain on the system itself. In doing this, the process will be tested on programs that are known to have these vulnerabilities, obtained from the website cwe.mitre.org. The testing will measure the time and accuracy of the process on different problem sizes and compare it to other processes that are known to be capable of solving this problem.