Scholarlink

As electric vehicles (EVs) rapidly gain global adoption, optimizing grids and analyzing charging behavior have become increasingly important. However, due to the sensitivity of charging data, ensuring privacy during data analysis has become a critical issue. This paper proposes a privacy-preserving kNN classification scheme based on FHE that ensures data privacy by performing classification over encrypted data. The scheme uses an optimized matrix method to compute Euclidean distances between encrypted data, significantly reducing computational complexity. Experimental evaluations on multiple publicly available datasets show that the scheme achieves high computational efficiency and classification accuracy while preserving privacy.

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 Fully Homomorphic Encryption-Based KNN Classification Scheme for Electric Vehicles Data

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Zhicheng Li,
  • Jian Xu,
  • Nan Zhang,
  • Teng Lu,
  • Peijun Li,
  • Nian Wang,
  • Qiuyue Wang

摘要

As electric vehicles (EVs) rapidly gain global adoption, optimizing grids and analyzing charging behavior have become increasingly important. However, due to the sensitivity of charging data, ensuring privacy during data analysis has become a critical issue. This paper proposes a privacy-preserving kNN classification scheme based on FHE that ensures data privacy by performing classification over encrypted data. The scheme uses an optimized matrix method to compute Euclidean distances between encrypted data, significantly reducing computational complexity. Experimental evaluations on multiple publicly available datasets show that the scheme achieves high computational efficiency and classification accuracy while preserving privacy.