Year: 2024
Type: book-chapter
Abstract: UAV Ad hoc Networks (UANETs) play a critical role in applications that necessitate secure and resilient communication, including data collection and surveillance.UANETs encounter substantial security challenges as a result of their decentralized and dynamic characteristics.One such challenge is the potential for malicious nodes to disrupt operations through the discarding of packets.The Drop Attack Detection Algorithm (DADA-UANET), which utilises supervised machine learning to improve network security, is the subject of this research.A combination of Logistic Regression (LR), Decision Tree (DT), and K-Nearest Neighbours (KNN) is utilised to differentiate between normal and malicious nodes efficiently.Our methodology incorporates an original implementation of linear regression to evaluate the credibility of nodes on a periodic basis by analysing their past actions.The experimental findings derived from a comparative analysis demonstrate that our approach attains an enhanced level of performance, surpassing established methodologies by as much as 92% in classification accuracy when LR and KNN are employed.The integration of DADA-UANET substantially enhances the robustness of unmanned aerial vehicle (UAV) communication in the face of advanced cyber threats, thereby guaranteeing more dependable functioning of vital applications.
Authors: Said Neciri, Noureddine Chaib, Chabane Djeddi
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4402009474