A Novel Topological Data Analysis-Enhanced Reputation System for UAV Networks

Document Type

Article

Publication Date

1-1-2026

Abstract

Uncrewed Aerial Vehicles (UAVs) have become indispensable components of the modern Internet of Things (IoT) ecosystem, increasingly used for delivery, transportation, inspection, and mapping. However, the reliability, security, and safety of UAVs are among the public's top concerns, as they operate in close proximity to one another and to other objects. This paper proposes an enhanced LIghtweight Blockchain-based REputation (LIBRE) system that integrates a novel Topological Data Analysis (TDA)-based reputation scoring mechanism to improve the reliability and performance of a UAV network by monitoring, tracking, and selecting the most appropriate agents to perform tasks. LIBRE captures intrinsic geometric and topological structures in UAV interaction data, enabling robust detection of anomalies and malicious behavior. Newly registered UAVs enter the network in a provisional limited-access state while trust evidence is accumulated from direct behavioral observations and peer feedback. The initial reputation value represents an unverified state rather than evidence of malicious behavior, and broader network privileges are granted only after sufficient behavioral evidence has been collected and the resulting reputation satisfies the trust threshold. Reputation values are recorded on a private blockchain to provide integrity, traceability, and auditability, while fine-grained behavioral data remain at the edge and only aggregated reputation information and cryptographic hashes are stored on-chain. A proof-of-concept prototype implemented on a private Ethereum Blockchain validated the effectiveness of the TDA-based LIBRE scheme, demonstrating significant improvements in threat detection and mitigation compared to traditional methods.

Publication Title

IEEE Transactions on Emerging Topics in Computing

Link to Full Text

Share

COinS