Research of Intrusion Protection System using correlation policy

Shuo Li, Quan Zhang2015 · DOI

Abstract

Traditional network security often suffers from the limitations of isolated technologies. This paper proposes an integrated Intrusion Protection System (IPS) that leverages a correlation policy to unify firewalls, Intrusion Detection Systems (IDS), and honeypots. By extending firewall redirection rules and implementing a custom linkage plug-in, the system achieves proactive defense against internal threats and "zero-day" attacks. The architecture ensures that suspicious traffic is not merely blocked but redirected to honeypots for analysis, effectively closing the "opportunity window" typically exploited by attackers.

The Fragmentation of Modern Network Security

Network security has historically relied on static, fragmented tools. Firewalls excel at perimeter defense but remain blind to internal lateral movement or backdoor connections. Intrusion Detection Systems (IDS) can identify malicious patterns but often require manual intervention, creating a critical time gap between detection and response. Honeypots provide deep insights into attacker behavior but are inherently passive.

In modern environments, attackers utilize zero-day exploits and internal "rebound" connections (Trojans) to bypass these individual layers. This research seeks to bridge these gaps by creating a cohesive system where these tools communicate and act in concert through automated redirection.

System Architecture and Methods

The core of the proposed system is a linkage plug-in module that facilitates real-time communication between three main components:

  1. Firewall (Netfilter/iptables): Acts as the central control point, managing dynamic redirection rules.
  2. IDS (Snort): Deployed on internal hosts and honeypots to sniff traffic and detect malicious signatures.
  3. Honeypots: Decoy systems (running both Windows and Linux) designed to attract and trap attackers for behavioral study.
The network topology of the proposed integrated IPS, showing the placement of firewalls, NIPS, and the DMZ area relative to internal hosts and honeypots.

Operational Workflow

When the IDS detects suspicious activity, it generates a "redirection struct" containing the source/destination IP and port information. This struct is sent to the firewall, which immediately updates its rules to shunt that specific traffic stream away from production assets and into a honeypot. This allows the system to "study" the attacker in a safe, isolated environment while protecting the actual network.

Flowchart of the response process for external data packets, illustrating the decision logic between rule consistency, Snort alarms, and honeypot redirection.

Experimental Evaluation

The researchers tested the system in a custom simulation environment featuring a gateway, external attack hosts, internal production hosts, and a virtualized honeynet. Three primary scenarios were evaluated:

  • External Attacks: During a simulated virus file transfer to an internal host, the system successfully redirected the malicious stream to a honeypot, leaving the target host unaffected.
  • Internal "Rebound" Attacks: A simulated Trojan on an internal host attempted to scan the network. The NIDS detected the outbound command-and-control flow, and the firewall redirected the traffic to a honeypot, neutralizing the threat.
  • Zero-Day Defense: While Snort initially missed an unknown exploit, the system triggered an alert once the attacker tampered with the honeypot. Data from the honeypot was then used to generate new rules, blocking the threat across the entire network.

Conclusion and Future Outlook

The integrated IPS demonstrates a significant improvement over standalone security tools by automating the response to complex threats. By reducing the "opportunity window" to near-zero, the system provides a robust defense against both known and unknown vulnerabilities.

However, the authors note that zero-day defense still requires some human intervention to verify honeypot logs before permanent rules are established. Additionally, future work must focus on honeypot isolation techniques to ensure that a compromised decoy cannot be used as a "springboard" for further internal attacks.

Summary updated

Abstract

Aimed at the limitations of single technology in network security defense, an Intrusion Protection System using correlation policy was proposed. This system consists of firewall, Intrusion Detection System and honeypot. The system extends the firewall redirection rules, sets up honeypot host monitors, and comes up with the linkage plug-in module of firewall, Intrusion Detection System and honeypot to achieve the close interaction of the three modules. On the basis of traditional Intrusion Protection System, the system based on proactive defense to defense the attack from internal and "0day attack" by honeypot.
BibTeX
@misc{li2015research,
  author = {Shuo Li and Quan Zhang},
  title = {{Research of Intrusion Protection System using correlation policy}},
  year = {2015},
  doi = {10.2991/meita-15.2015.117},
}
Figure