Virtual Resource Management for Cell Broadcast Alert Handling

This paper introduces an autonomous agent-based system for intelligent redirecting Cell Broadcast (CB) messages to e-mail using a Raspberry Pi 4B (RPi 4B) connected to a modem, addressing the challenge of limited access to emergency alerts. The concept uses a virtual resource that can act alone upon the occurrence of a stimulus and which manages the dissemination of critical public safety information accurately and in a timely manner. The approach combines an affordable computing platform with an Artificial Intelligence (AI) model that acts as a message filter. The RPi 4B monitors the modem via AT commands, allowing the AI module to classify transmissions and forward alerts to a designated e-mail address. Unlike existing academic works, which primarily focus on weather or seismic triggers, which do not provide equivalent ready-made functionality, the proposed system is designed to expand the system’s alerting capabilities worldwide. Key contributions include expanding access to CB messages, demonstrating a cost-efficient AI-enhanced architecture and delivering a deployable, user-friendly system. Experimental evaluation indicates redirection delays between 57 and 70 seconds, dominated by the modem acquisition stage, while the internal processing latency, from message reception to database storage, was between 5 and 10 seconds.