Scholarlink

<p>The real-time communication between vehicles and roadside infrastructure, vehicular ad hoc networks (VANETs) are essential to intelligent transport networks. It is still difficult to ensure safe communications and information security because of changing digital dangers and flexible networks. The aims to develop an adaptive and robust mechanism for detecting false data injection and bogus report attacks in VANET environments. A hybrid detection framework, termed Fine-Tuned Tabu Search–based Isolation Forest (FTS-IF), is proposed. The method optimizes key Isolation Forest hyperparameters using Fine-Tuned Tabu Search to improve anomaly detection performance. Experiments are conducted on a pre-processed dataset generated from 100 randomly deployed sensor nodes within a 100 × 100&#xa0;m² area, with Min–Max normalization applied. The proposed FTS-IF model achieves an accuracy of 89%, precision of 88%, recall of 80%, and an F1-score of 82%, outperforming conventional machine learning-based detection methods. The results demonstrate that FTS-IF effectively distinguishes legitimate data from false reports, enhancing cybersecurity and secure data management in VANETs. The proposed framework offers an efficient and practical solution for detecting malicious node behavior and false data injection attacks in VANETs.</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

Machine Learning Method for False Data Detection and Secure Communication in VANETs

  • 期刊论文

  • 发表日期：

    2026年2月3日

查看全文
  • Veena S. Badiger,
  • Debanjan Ghosh,
  • Vikas Sagar,
  • B. Pakruddin,
  • Santosh Kumar Behera

摘要

The real-time communication between vehicles and roadside infrastructure, vehicular ad hoc networks (VANETs) are essential to intelligent transport networks. It is still difficult to ensure safe communications and information security because of changing digital dangers and flexible networks. The aims to develop an adaptive and robust mechanism for detecting false data injection and bogus report attacks in VANET environments. A hybrid detection framework, termed Fine-Tuned Tabu Search–based Isolation Forest (FTS-IF), is proposed. The method optimizes key Isolation Forest hyperparameters using Fine-Tuned Tabu Search to improve anomaly detection performance. Experiments are conducted on a pre-processed dataset generated from 100 randomly deployed sensor nodes within a 100 × 100 m² area, with Min–Max normalization applied. The proposed FTS-IF model achieves an accuracy of 89%, precision of 88%, recall of 80%, and an F1-score of 82%, outperforming conventional machine learning-based detection methods. The results demonstrate that FTS-IF effectively distinguishes legitimate data from false reports, enhancing cybersecurity and secure data management in VANETs. The proposed framework offers an efficient and practical solution for detecting malicious node behavior and false data injection attacks in VANETs.