IPMES: A Tool for Incremental TTP Detection Over the System Audit Event Stream
Hongwei Li, P.-L. Liu, Bo-Wei Lin, Yi‐Chun Liao, Yennun Huang · 2024
Advanced persistent threat (APT) cyberattacks are serious threats to corporations and governments. The prolong dwell time associated with APTs significantly increase the difficulty on detecting them in provenance graphs. To reduce the detection complexity, some works have demonstrated the effectiveness of employing pattern matching on provenance graphs in conjunction with APT lifecycle models to pinpoint short-duration attack steps, also known as “tactics, techniques, and procedures” (TTPs). However, when dealing with more complex TTPs, particularly those involving graph-based and partial ordering, few tools can incrementally and efficiently handle them. In this paper, we present IPMES11https://github.com/littleponywork/IPMES, a tool that has been publicly released to address this gap. By leveraging specific optimizations, it provides efficient incremental matching for those TTPs, and can handle practical system audit event streams. Experiments conducted on synthetic and real-world data demonstrated the practical feasibility of IPMES in TTP detection.