Framework de IA para Detecção e Mitigação de Ameaças em Redes Móveis Privativas

Private mobile networks require security mechanisms capable of operating in distributed environments. This paper proposes an AI-based framework for threat detection and mitigation in an architecture with a remote radio access network (RAN) over Starlink. The proposed solution integrates security tools such as Wazuh and Suricata, along with Python scripts, the N8N automation platform, and large language models to correlate events and support semi-automated responses, including blocking suspicious IP addresses and sending notifications. In a controlled denial-of-service scenario, the system correlated 28 alerts within 24 hours and applied containment actions within up to 1 minute, demonstrating the operational feasibility of the proposed approach.