Trust-based routing methodology in uav swarm networks based on traffic analysis and anomaly detection

Abstract

Subject matter. the subject of the research is the process of ensuring secure routing and data exchange among unmanned aerial vehicles (UAVs) within swarm networks under cyber threat conditions, particularly Black Hole-type attacks. Goal. The purpose of this study is to develop and simulate a secure information transmission mechanism for UAV swarm networks that takes into account node trust levels and enables the identification of malicious participants based on behavioral analysis. The study also aims to establish a methodology for secure and energy-efficient routing in FANETs based on blockchain technologies and trust evaluation models, ensuring cyber resilience, data integrity, and minimal resource usage. Tasks the following objectives were addressed during the research. Conduct a comprehensive analysis of vulnerabilities in traditional routing protocols used in FANETs to identify potential threats to information security: justify the use of trust-based mechanisms to improve routing resilience against internal attacks; implement a Black Hole attack model within the NS-3 simulation environment to analyze its impact on swarm network performance; develop a mechanism for counting forwarded packets per node as a foundation for a trust evaluation system among agents; visualize simulation results to support analysis and comparison of proposed methods. Methods: the research employs simulation modeling of FANETs in NS-3.36, using the RandomWaypoint mobility model and the AODV routing protocol. Methods include statistical analysis of packet forwarding metrics and graphical representation of trust metrics. The developed code simulates the Black Hole attack by manipulating the NetDevice layer and logs all transmitted packets in CSV format for post-processing. A combination of simulation tools, analytical analysis, and visualization techniques was applied to evaluate system performance under dynamic conditions. Results. The results demonstrate the effectiveness of the proposed approach in detecting malicious nodes within the swarm network. Trust metrics revealed anomalous attacker behavior, such as the absence of packet forwarding, distinguishing them from normal nodes. This allows for timely identification and exclusion of threats from the routing process. Graphical visualization clearly displays node activity distribution, facilitating result interpretation without the need for in-depth log analysis. Conclusions. The proposed trust-based mechanism, combined with node activity analysis, effectively protects FANET networks against Black Hole attacks. Future improvements may include integrating more advanced trust assessment methods, such as multifactor analysis, blockchain, or machine learning. Developing adaptive routing algorithms capable of autonomously isolating or excluding suspicious nodes is also recommended