面向无人机网络的多类分类增强入侵检测系统
摘要
无人机 (UAV) 由于涌现 6G 系统和网络的出现，在 various applications 中变得越来越流行。然而，its 广泛采用也引发了关于 security vulnerabilities 的关注，使得 develop 可靠的 intrusion detection system (IDS) 对确保 UAVs safety 和 mission success 至关重要。本文提出一种 new IDS for UAV networks。采用 binary-tuple representation 对 class labels 进行编码，并 employed deep learning-based approach 用于 classification。该 proposed system 通过 capture complex class relationships 和 temporal network patterns 来 enhance intrusion detection。此外，对 common features of different UAVs 进行 cross-correlation study，以 discard potentially 可能误导该 IDS 的 classification 的 correlated features。full study 采用 UAV-IDS-2020 dataset 进行，we assessed the performance of the proposed IDS using different evaluation metrics。experimental results 突出了 proposed multiclass classifier model 在 accuracy 上达到 95% 的效果。
关键词
引用
@article{arxiv.2406.10417,
title = {Enhanced Intrusion Detection System for Multiclass Classification in UAV Networks},
author = {Safaa Menssouri and Mamady Delamou and Khalil Ibrahimi and El Mehdi Amhoud},
journal= {arXiv preprint arXiv:2406.10417},
year = {2025}
}