Scholarlink

<p>In the thriving era of the modern transportation system, electric vehicles (EVs) have gained massive popularity due to numerous privileges such as being ecosystem-friendly and economically cheaper than conventional transportation modes. Therefore, the issues associated with EVs have also increased drastically. When EVs are connected to charging stations (CSs), they result in the generation of data that poses notable privacy challenges. This data contains sensitive information like charging status, electrical operational range, vehicle ID, and zip code. The traditional privacy preservation approaches are not effective in protecting the privacy of the whole event of a data stream dynamically. Hence, this study proposes a dynamic approach, Event-Wise Differential Privacy (EWDP), which aims to protect the privacy of EVs and charging stations (CSs). This research has been classified into three parts. Firstly, the sampling points or events are identified. In the second part, these events are selected based on the Novel Probabilistic Approach (NPA), this approach will then select a certain number of events to proceed further. In the third step, Laplacian noise is added in the selected events using suitable utility parameters to protect the multiclasses of the user’s sensitive information. This proposed technique has mitigated the risk of re-identification vector attacks, proved useful for protecting the privacy of the user, and achieved an optimum balance between data utility and privacy. The experimental evaluations demonstrate a fully functional practical approach of EWDP using the NPA, followed by the computing of Mean Average Error (MAE), Absolute Error at different privacy budgets <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\({\epsilon }\)</EquationSource> </InlineEquation> and the computational efficiency for our proposed technique.</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

EWDP: event-wise differential privacy for efficient electric vehicles infrastructure

  • 期刊论文

  • 发表日期：

    2025年10月18日

查看全文
  • Mohsin Ali,
  • Muneeb Ul Hassan,
  • Pei-Wei Tsai,
  • Jinjun Chen

摘要

In the thriving era of the modern transportation system, electric vehicles (EVs) have gained massive popularity due to numerous privileges such as being ecosystem-friendly and economically cheaper than conventional transportation modes. Therefore, the issues associated with EVs have also increased drastically. When EVs are connected to charging stations (CSs), they result in the generation of data that poses notable privacy challenges. This data contains sensitive information like charging status, electrical operational range, vehicle ID, and zip code. The traditional privacy preservation approaches are not effective in protecting the privacy of the whole event of a data stream dynamically. Hence, this study proposes a dynamic approach, Event-Wise Differential Privacy (EWDP), which aims to protect the privacy of EVs and charging stations (CSs). This research has been classified into three parts. Firstly, the sampling points or events are identified. In the second part, these events are selected based on the Novel Probabilistic Approach (NPA), this approach will then select a certain number of events to proceed further. In the third step, Laplacian noise is added in the selected events using suitable utility parameters to protect the multiclasses of the user’s sensitive information. This proposed technique has mitigated the risk of re-identification vector attacks, proved useful for protecting the privacy of the user, and achieved an optimum balance between data utility and privacy. The experimental evaluations demonstrate a fully functional practical approach of EWDP using the NPA, followed by the computing of Mean Average Error (MAE), Absolute Error at different privacy budgets \({\epsilon }\) and the computational efficiency for our proposed technique.