Applications and Challenges of Flying Ad-Hoc Networks in Disaster Response: A Review

Abstract

Natural disasters cause massive destruction to power infrastructure and communication networks and limit access to proper emergency response. Unmanned aerial vehicles (UAV)-based Flying ad hoc networks (FANETs) offer a useful communication layer through aerial observation, line-of-sight relay support, and cooperative multi-hop networking to restore connections in affected areas. However, the available FANET literature uses different assumptions about mobility, channel conditions, interference, traffic profiles, and performance metrics, which limits proper cross-study comparison and protocol selection for emergency use. In this review, we examine the most recent FANET architectures and disaster-response protocols, including medium access control, routing, topology management, and energy-aware operation. We also analyze connectivity-preservation strategies under rapid topology change, including mobility prediction, link-quality estimation, geographic relay selection, and control-plane design. We then evaluate the role of multi-agent reinforcement learning (MARL) in decentralized task allocation, relay selection, spectrum access, and power control under partial observability. Future research should prioritize reproducible experimental evaluation of FANET architectures, MAC protocols, routing strategies, and AI-enabled coordination under realistic disaster conditions.