Abstract : It is argued that runtime verification techniques can be used to identify unknown application security vulnerabilities that are a consequence of unexpected execution paths in software. A methodology is proposed that can be used to build a model of expected application execution paths during the software development cycle. This model is used at runtime to detect exploitation of unknown security vulnerabilities using anomaly detection style techniques. The approach is evaluated by considering its effectiveness in identifying 19 vulnerabilities across 26 versions of Apache Struts over a 5 year period.
https://hal.inria.fr/hal-01633685 Contributor : Hal IfipConnect in order to contact the contributor Submitted on : Monday, November 13, 2017 - 11:46:44 AM Last modification on : Monday, June 15, 2020 - 1:38:03 PM Long-term archiving on: : Wednesday, February 14, 2018 - 12:58:32 PM
Olgierd Pieczul, Simon N. Foley. Runtime Detection of Zero-Day Vulnerability Exploits in Contemporary Software Systems. 30th IFIP Annual Conference on Data and Applications Security and Privacy (DBSec), Jul 2016, Trento, Italy. pp.347-363, ⟨10.1007/978-3-319-41483-6_24⟩. ⟨hal-01633685⟩