Scholarlink

<p>Vehicular Ad-Hoc Networks (VANETs) are crucial for managing traffic flow and enhancing safety in Intelligent transportation systems (ITS) through efficient data transmission among vehicular nodes. However, VANETs face significant challenges, including susceptibility to malicious activities, complex routing issues, dynamic sensor activities, and unpredictable failures in vehicles, which can compromise network security and performance. This study proposes an optimized multi-objective algorithm to address these challenges by improving cluster head selection and ensuring secure data transmission. The algorithm employs a multi-objective function for effective cluster head selection and uses a modified enhanced station-to-station (e-STS) cryptographic algorithm for secure communication with roadside units (RSUs). In cases of detected malicious activity, the protocol incorporates Isolation forest and Autoencoder techniques for attack classification, supported by optimized feature selection. The proposed system adapts to the dynamic topology changes of VANETs and enhances network performance metrics, including recall (0.96%), accuracy (0.985%), F1 score (0.91%), and precision (0.872%). Experimental results demonstrate improvements in cluster formation (10 clusters), packet delivery ratio (93%), cluster efficiency (95%), throughput (10199 kbps), and delay (0.31s). The performance of the proposed method is benchmarked against recent algorithms such as FASMO-DMN (Kaur, Kakkar Ad Hoc Netw 136:102961 <CitationRef CitationID="CR1">2022</CitationRef>), MetaLearn (Nahar, Das Ad Hoc Netw 138:102996 <CitationRef CitationID="CR2">2023</CitationRef>), and ACO-K-means (Kadam, Vaze, Todmal Wireless Pers Commun 132(1):305–328 <CitationRef CitationID="CR3">2023</CitationRef>), showing superior results in key performance indicators using MATLAB and SUMO simulators. This research contributes to advancing VANET security and mobility, addressing existing limitations, and improving overall ITS functionality.</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

A multi-objective approach for secure cluster based routing & attack classification in VANETs

  • 期刊论文

  • 发表日期：

    2025年3月11日

查看全文
  • Aradhana Behura,
  • Arun Kumar,
  • Puneet Kumar Jain

摘要

Vehicular Ad-Hoc Networks (VANETs) are crucial for managing traffic flow and enhancing safety in Intelligent transportation systems (ITS) through efficient data transmission among vehicular nodes. However, VANETs face significant challenges, including susceptibility to malicious activities, complex routing issues, dynamic sensor activities, and unpredictable failures in vehicles, which can compromise network security and performance. This study proposes an optimized multi-objective algorithm to address these challenges by improving cluster head selection and ensuring secure data transmission. The algorithm employs a multi-objective function for effective cluster head selection and uses a modified enhanced station-to-station (e-STS) cryptographic algorithm for secure communication with roadside units (RSUs). In cases of detected malicious activity, the protocol incorporates Isolation forest and Autoencoder techniques for attack classification, supported by optimized feature selection. The proposed system adapts to the dynamic topology changes of VANETs and enhances network performance metrics, including recall (0.96%), accuracy (0.985%), F1 score (0.91%), and precision (0.872%). Experimental results demonstrate improvements in cluster formation (10 clusters), packet delivery ratio (93%), cluster efficiency (95%), throughput (10199 kbps), and delay (0.31s). The performance of the proposed method is benchmarked against recent algorithms such as FASMO-DMN (Kaur, Kakkar Ad Hoc Netw 136:102961 2022), MetaLearn (Nahar, Das Ad Hoc Netw 138:102996 2023), and ACO-K-means (Kadam, Vaze, Todmal Wireless Pers Commun 132(1):305–328 2023), showing superior results in key performance indicators using MATLAB and SUMO simulators. This research contributes to advancing VANET security and mobility, addressing existing limitations, and improving overall ITS functionality.