With the ongoing involvement of computers in all facets of modern society, both private and public organizations increasingly rely on computer networks to support key portions of their critical infrastructure. Consequently, those networks contain increasingly more sensitive information, making them prime targets for would-be attackers. The goal of a typical intrusion detection system (IDS) is to protect such a network by identifying attempted attacks, alerting administrators of malicious behavior, and defending against known exploits where possible. If a previously unknown exploit is detected, it is often left to the network administrator to manually thwart the attack and protect any uncompromised nodes. Unfortunately, this task can often be tedious and intractable as it may require a deep understanding of the exploited vulnerability and how to correct the underlying system. Pikewerks proposes a high-performance solution capable of detecting and adapting to this increased level of threat. This proposed system will advance state-of-the-art IDS research by functioning as a fully-automated, instruction-level, host-based intrusion detection and prevention system designed to support self-healing, self-hardening network architectures against zero-day exploits.
Keywords: Vulnerability Signatures, Intrusion Detection System (Ids), Self-Healing, Automatic Signature Generation, Headless Administration
