FURI | Summer 2025

Using Petri Nets to Model Subtle Network Interactions to Improve Intrusion Detection

Current cybersecurity systems and modern firewalls rely heavily on observing network flows passively to detect threats. However, the attacks have evolved to exploit subtle packet behaviors, making the systems vulnerable to newer evasion techniques.

This research aims to use Petri nets to model the dynamic interactions between network packets, firewalls and VPNs, looking at “byte patterns” to capture the minute details or anomalies, which would otherwise be overlooked by traditional tools.

Modeling interaction this way could significantly improve our understanding of infection mechanisms of how malware bypasses security, leading to enhanced intrusion detection effectiveness and overall cybersecurity defenses.