Scholarlink

<p>Vehicular Ad Hoc Networks (VANETs) play a crucial role in intelligent transportation systems by enabling real-time communication among vehicles and roadside units. However, their decentralized architecture, high mobility, and the limited resources of their computing units make them very vulnerable to cybersecurity threats such as data tampering, spoofing or impersonation, and denial of service. The traditional security and threat detection strategies and procedures generally cannot adapt to the dynamic, heterogeneous nature of the VANET environment, which results in inefficient data treatment and poor intrusion resilience. It proposes a new ARIMA-SK-EELM called Autoregressive Integrated Moving Average with Stable Kernel-Enhanced Extreme Learning Machine-based prediction and data fusion model with the Threefish Algorithm (TFA) to secure their communication. The ARIMA model is used to predict data trends for a number of reasons. It forecasts, based on a limited number of sensory inputs, to minimize unnecessary transmissions. The SK-EELM refines predictions to achieve high accuracy and robustness to interference, and the TFA detects wearable equipment and ensures encryption and data integrity during transmission. Simulation results show that the ARIMA-SK-EELM approach, with parameters ζ = 0.02, λ = 0.0009, and β = 0.7, achieves better accuracy, minimal mean error, and a high forecast success rate. Statistical analysis using Paired t-test shows the proposed model predicts more accurately than baselines. The proposed approach offers a comprehensive solution for intelligent threat detection and secure data management in context-relevant and modular form, to minimize theft detection and dissemination of data for use in active 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

Intelligent Threat Detection and Secure Data Management in VANETs Using Machine Learning

  • 期刊论文

  • 发表日期：

    2026年2月5日

查看全文
  • Aarsi Kumari,
  • Ritesh Rastogi,
  • K. Prakash,
  • Satya Ranjan Das,
  • N. Vasantha Kumari

摘要

Vehicular Ad Hoc Networks (VANETs) play a crucial role in intelligent transportation systems by enabling real-time communication among vehicles and roadside units. However, their decentralized architecture, high mobility, and the limited resources of their computing units make them very vulnerable to cybersecurity threats such as data tampering, spoofing or impersonation, and denial of service. The traditional security and threat detection strategies and procedures generally cannot adapt to the dynamic, heterogeneous nature of the VANET environment, which results in inefficient data treatment and poor intrusion resilience. It proposes a new ARIMA-SK-EELM called Autoregressive Integrated Moving Average with Stable Kernel-Enhanced Extreme Learning Machine-based prediction and data fusion model with the Threefish Algorithm (TFA) to secure their communication. The ARIMA model is used to predict data trends for a number of reasons. It forecasts, based on a limited number of sensory inputs, to minimize unnecessary transmissions. The SK-EELM refines predictions to achieve high accuracy and robustness to interference, and the TFA detects wearable equipment and ensures encryption and data integrity during transmission. Simulation results show that the ARIMA-SK-EELM approach, with parameters ζ = 0.02, λ = 0.0009, and β = 0.7, achieves better accuracy, minimal mean error, and a high forecast success rate. Statistical analysis using Paired t-test shows the proposed model predicts more accurately than baselines. The proposed approach offers a comprehensive solution for intelligent threat detection and secure data management in context-relevant and modular form, to minimize theft detection and dissemination of data for use in active VANET environments.