Scholarlink

<p>The vehicle obtains the related information through VANET dissemination. However, the lack of consideration of visibility &#xa0;perspective and real-time circumstances in dissemination protocols development &#xa0;has a significant impact on the programs aimed at preventing accidents and ensuring human safety. In order to address the challenge of information dissemination based on dynamic visibility conditions, a cluster-based approach PV-DVMC is employed to effectively capture and model dynamism in dissemination according to current visibility when considering the mobility of vehicles. Proposed PV-DVMC is a novel method that aims to offer a dynamic dissemination strategy for varying and unpredictable real-time visibility conditions encountered by drivers and vehicles during motion. Additionally, we deploy the layering method in the cluster to reduce complexity and communication overhead. We further proposed DPLD; the predictive machine-learning model on&#xa0;a selected cluster to predict the layers based on classification and regression decision-tree approach for multiple visibility factors. PV-DVMC efficiency was&#xa0;assessed through simulations. In comparing other standard cluster based strategy without visibility perspective performance of&#xa0;the proposed is improved by reducing delay and collision ratio to 46.23% and 38.64% respectively, while the ratio of delivery of messages increases by 42.82%. Further the nuScene Dataset is &#xa0;utilized for dynamic visibility scenarios to evaluate the accuracy, precision, recall and F1-score performance of our proposed method.</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

PV-DVMC: a novel path visibility based reliable data dissemination in VANETs using machine learning with layered clustering

  • 期刊论文

  • 发表日期：

    2025年9月11日

查看全文
  • Deepak Gupta,
  • Shikha Gupta

摘要

The vehicle obtains the related information through VANET dissemination. However, the lack of consideration of visibility  perspective and real-time circumstances in dissemination protocols development  has a significant impact on the programs aimed at preventing accidents and ensuring human safety. In order to address the challenge of information dissemination based on dynamic visibility conditions, a cluster-based approach PV-DVMC is employed to effectively capture and model dynamism in dissemination according to current visibility when considering the mobility of vehicles. Proposed PV-DVMC is a novel method that aims to offer a dynamic dissemination strategy for varying and unpredictable real-time visibility conditions encountered by drivers and vehicles during motion. Additionally, we deploy the layering method in the cluster to reduce complexity and communication overhead. We further proposed DPLD; the predictive machine-learning model on a selected cluster to predict the layers based on classification and regression decision-tree approach for multiple visibility factors. PV-DVMC efficiency was assessed through simulations. In comparing other standard cluster based strategy without visibility perspective performance of the proposed is improved by reducing delay and collision ratio to 46.23% and 38.64% respectively, while the ratio of delivery of messages increases by 42.82%. Further the nuScene Dataset is  utilized for dynamic visibility scenarios to evaluate the accuracy, precision, recall and F1-score performance of our proposed method.