SAGE: Intrusion Alert-driven Attack Graph Extractor (short paper)

Azqa Nadeem, Sicco Verwer, Shanchieh Jay Yang

View presentation: 2021-10-27T16:00:00Z GMT-0600 Change your timezone on the schedule page
2021-10-27T16:00:00Z
Exemplar figure, described by caption below
An alert-driven attack graph shows attacker strategies extracted from actions observed through intrusion alerts. Each graph shows the strategies of all attackers that obtain an objective on a particular victim. The image shows an attack graph of data_manipulation over remoteware-cl, where 3 teams successfully exploit it: Team 5 exploits it twice, while Teams 1 and 8 exploit it once. The S-PDFA identifies three ways of exploiting the objective based on the actions that lead up to it. Teams 5 and 8 share a significant portion of a strategy, while Team 1 has found the shortest path to reach the objective.
Fast forward

Direct link to video on YouTube: https://youtu.be/TxUye31Eqjs

Abstract

Attack graphs (AG) are used to assess pathways availed by cyber adversaries to penetrate a network. State-of-the-art approaches for AG generation focus mostly on deriving dependencies between system vulnerabilities based on network scans and expert knowledge. In real-world operations however, it is costly and ineffective to rely on constant vulnerability scanning and expert-crafted AGs. We propose to automatically learn AGs based on actions observed through intrusion alerts, without prior expert knowledge. Specifically, we develop an unsupervised sequence learning system, SAGE, that leverages the temporal and probabilistic dependence between alerts in a suffix-based probabilistic deterministic finite automaton(S-PDFA) – a model that accentuates infrequent severe alerts and summarizes paths leading to them. AGs are then derived from the S-PDFA. Tested with intrusion alerts collected through Collegiate Penetration Testing Competition, SAGE compresses several thousands of alerts into only a handful of AGs. These AGs reflect the strategies used by participating teams. The resulting AGs are succinct, interpretable, and enable analysts to derive actionable insights, e.g., attackers tend to follow shorter paths after they have discovered a longer one.