AI-Powered Decentralized Resilience: Mesh Networks for Disaster Communication in Panama

Abstract

The objective of this research is to design and evaluate an optimized Mesh network model with self-healing protocols assisted by Artificial Intelligence (IA), to ensure resilient communications during disasters in Panama. The methodology employed an experimental design based on a dynamic simulation environment (Constelación Mesh), where a Markov Chain model was implemented for network state prediction and a reinforcement learning algorithm (MDP) for autonomous routing and resource allocation decision-making. The results demonstrate that the AI model predicts node failures with an accuracy of over 90% and reconfigures data routes in milliseconds, maintaining communication integrity. Furthermore, the AI’s autonomous actions successfully prevented network congestion and reduced critical node downtime by 75% compared to a model without AI. It is concluded that the integration of AI into Mesh networks offers a viable and significantly more efficient solution for emergency communications in the Panamanian context, representing a crucial advancement for disaster management and the resilience of the country’s critical infrastructure.