Scholarlink

<p>Vehicular Ad Hoc Networks (VANETs) enable vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications, making them vulnerable to cyberattacks such as unauthorized access, message tampering, and denial-of-service attacks. To report these limitations, this research proposes a Cluster-Aware Swarm-Intelligent Principal Component Analysis (CAS-PCA)–based intrusion detection framework for VANET cybersecurity. The framework organizes vehicles into dynamic clusters using K-Means clustering, extracts discriminative traffic and behavioral features through Principal Component Analysis (PCA), and optimizes intrusion detection thresholds and cluster-head selection using particle swarm optimization (PSO). Simulation results demonstrate that CAS-PCA improves both network efficiency and security performance. In addition to reduced delay (47.53) and request failure (0.06), the proposed method achieves a detection accuracy of 96.4%, a FPR of 2.6%, and a detection latency of 63.5&#xa0;ms, outperforming baseline and hybrid intrusion detection system (IDS) approaches. These results confirm that CAS-PCA provides an efficient and reliable intrusion detection solution for dynamic VANET environments.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Cluster-Aware Intrusion Detection Framework for Enhancing Cyber Security in VANETs

  • 期刊论文

  • 发表日期：

    2026年4月13日

查看全文
  • Syed Rashid Anwar,
  • Anamika Chaudhary,
  • M. Sudharsan,
  • Bichitrananda Patra,
  • Anitha D Souza J

摘要

Vehicular Ad Hoc Networks (VANETs) enable vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications, making them vulnerable to cyberattacks such as unauthorized access, message tampering, and denial-of-service attacks. To report these limitations, this research proposes a Cluster-Aware Swarm-Intelligent Principal Component Analysis (CAS-PCA)–based intrusion detection framework for VANET cybersecurity. The framework organizes vehicles into dynamic clusters using K-Means clustering, extracts discriminative traffic and behavioral features through Principal Component Analysis (PCA), and optimizes intrusion detection thresholds and cluster-head selection using particle swarm optimization (PSO). Simulation results demonstrate that CAS-PCA improves both network efficiency and security performance. In addition to reduced delay (47.53) and request failure (0.06), the proposed method achieves a detection accuracy of 96.4%, a FPR of 2.6%, and a detection latency of 63.5 ms, outperforming baseline and hybrid intrusion detection system (IDS) approaches. These results confirm that CAS-PCA provides an efficient and reliable intrusion detection solution for dynamic VANET environments.