Scholarlink

<p>As vehicular communication becomes increasingly integral to intelligent transportation systems, ensuring robust security in Vehicular Ad-Hoc Networks (VANETs) is critical. While most studies focus on vulnerabilities within the CSMA/CA protocol, this paper explores the TDMA protocol, specifically Distributed Time Division Multiple Access (DTMAC). We identify four novel types of exploitable greedy behaviours. To detect these behaviours, we propose a watchdog model designed to analyse network traffic, extract relevant features, and generate datasets at varying levels of network density. This proposed solution employs the Grid Search Cross-Validation (GSCV) exhaustive parameter search technique, combined with the Radial Basis Function Kernel of the Support Vector Machine (RBF-SVM) algorithm. A comparison with other machine learning (ML) algorithms was conducted based on key performance metrics. Experimental simulations assessing the solution’s ability to counter greedy behaviour attack demonstrated that the proposed approach achieved an overall accuracy of 95% in the low-density scenario and 80% in the high-density scenario.</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

An RBF-SVM kernel-powered framework for enhancing the detection of greedy behaviour attacks in vehicular Ad Hoc networks

  • 期刊论文

  • 发表日期：

    2026年2月23日

查看全文
  • Tayssir Ismail,
  • Mohamed Hadded,
  • Nasreddine Hajlaoui,
  • Haifa Touati,
  • Samia Bouzefrane,
  • Paul Muhlethaler,
  • Leila Azouz Saidane

摘要

As vehicular communication becomes increasingly integral to intelligent transportation systems, ensuring robust security in Vehicular Ad-Hoc Networks (VANETs) is critical. While most studies focus on vulnerabilities within the CSMA/CA protocol, this paper explores the TDMA protocol, specifically Distributed Time Division Multiple Access (DTMAC). We identify four novel types of exploitable greedy behaviours. To detect these behaviours, we propose a watchdog model designed to analyse network traffic, extract relevant features, and generate datasets at varying levels of network density. This proposed solution employs the Grid Search Cross-Validation (GSCV) exhaustive parameter search technique, combined with the Radial Basis Function Kernel of the Support Vector Machine (RBF-SVM) algorithm. A comparison with other machine learning (ML) algorithms was conducted based on key performance metrics. Experimental simulations assessing the solution’s ability to counter greedy behaviour attack demonstrated that the proposed approach achieved an overall accuracy of 95% in the low-density scenario and 80% in the high-density scenario.