Secure Multi-Agent Artificial Intelligence Framework for Autonomous Drone Swarm Coordination in Disaster Response

Abstract

Autonomous drone swarms are increasingly being deployed for disaster response operations, including search and rescue, damage assessment, and emergency communication. Coordinating multiple autonomous drones in dynamic environments while ensuring secure communication remains a challenging problem. This paper proposes a secure multi-agent artificial intelligence framework that combines Multi-Agent Reinforcement Learning (MARL), edge computing, and blockchain technology to enable collaborative drone swarm coordination. Reinforcement learning agents dynamically optimize flight paths, task allocation, and obstacle avoidance, while blockchain-based authentication mechanisms ensure secure communication and prevent malicious interference. Extensive simulations demonstrate improved mission completion time, communication reliability, and operational efficiency compared with traditional centralized drone coordination approaches. The proposed framework provides a scalable and resilient solution for intelligent disaster management and emergency response systems.