TITLE:
A Cyber-Resilient UAV Swarm Framework for Fire-Fighting with AI-Based In-Flight Defect Inspection
AUTHORS:
Ahad Alotaibi, Abdullah Alrasheedi
KEYWORDS:
UAV Swarm, Fire-Fighting Missions, Cyber-Resilient UAV Systems, AI-Enabled In-Flight Defect Detection, UAV Cybersecurity, MITM Attack Mitigation, Secure Routing, GNS3 Simulation, Amazon Rekognition
JOURNAL NAME:
Journal of Computer and Communications,
Vol.14 No.1,
January
22,
2026
ABSTRACT: Drone swarms are increasingly deployed in fire-fighting missions due to their scalability, adaptability, and ability to operate in hazardous and dynamic environments. However, ensuring mission continuity in such safety-critical scenarios requires not only reliable drone performance but also strong protection against cyber threats targeting inter-drone communication and coordination. This paper presents a cyber-resilient UAV swarm framework for fire-fighting missions that integrates AI-enabled in-flight defect and integrity inspection with secure communication mechanisms. The proposed architecture employs a hierarchical formation in which a dedicated inspector drone periodically captures visual data of neighboring operational drones during the mission to detect structural defects, abnormal behavior, or integrity violations using AI-based visual analysis implemented with Amazon Rekognition. In parallel, cyber resilience is reinforced through subnet segmentation and Route Optimization for Autonomous Systems (ROAS), which mitigates Man-in-the-Middle (MITM) and traffic manipulation attacks. The framework is evaluated using a cyberattack simulation environment built in GNS3 with Ettercap for adversarial traffic injection. Experimental results demonstrate reduced attack success rates, early detection of drone defects, and improved operational reliability, ensuring secure and continuous swarm operation during fire-fighting missions. The proposed dual-layer approach provides a secure, self-verifying UAV swarm architecture suitable for safety-critical fire-fighting and emergency response applications.