AI-Powered Dual-Layer Network Security System Using Raspberry Pi

The rapid expansion of internet-connected devices has significantly increased exposure to cyber threats, particularly in small and medium-scale environments where enterprise-grade security solutions remain financially inaccessible. Traditional firewall and signature-based intrusion detection systems (IDS) are effective against known threats but lack adaptability to zero-day and behavioral attacks. This paper presents an AI-Powered Dual-Layer Network Security System implemented on Raspberry Pi 4, integrating DNS-level filtering using Pi-hole and packet-level intrusion detection using Snort3. In addition to signature-based monitoring, a lightweight AI-driven threat scoring and anomaly evaluation module is introduced to dynamically assess network risk levels. The system performs real-time log parsing, behavioral feature extraction, and automated IP blocking based on computed threat thresholds. The proposed architecture demonstrates that layered security with adaptive intelligence can be deployed on low-cost hardware while maintaining effective detection and prevention capabilities. Experimental observations validate improved responsiveness compared to standalone firewall-based systems.