摘要
针对智慧城市电表数据采集过程中面临用户隐私泄露的风险问题,提出了一种基于随机森林算法和聚合签名方法相结合的数据加密方法。利用随机森林方法提取用户用电特征信息作为判断用户是否具有窃电行为的判据。采用一次性假名算法、聚合签名方法与同态加密算法相结合的方法对传输的数据进行加密。最后,算例结果表明与传统的K-NN、SVM方法相比,所提随机森林方法对用户用电特征的分类准确度提升2.2%与4.45%,此外隐私保护的效果更好。
In view of the risk of user privacy leakage in the process of smart city meter data collection,a data encryption method based on the combination of random forest algorithm and aggregate signature method is proposed.Using the random forest method to extract the characteristic information of users'electricity consumption as the criterion to judge whether users have the behavior of stealing electricity.The method of one-time kana algorithm,aggregate signature method and homomorphic encryption algorithm is used to encrypt the transmitted data.Finally,the results show that compared with the traditional k-NN and SVM methods,the proposed random forest method improves the classification accuracy of user electricity characteristics by 2.2%and 4.45%,and the effect of privacy protection is better.
出处
《科技视界》
2020年第5期133-136,共4页
Science & Technology Vision
关键词
智慧城市
物联网
电能表
随机森林
同态加密
Smart city
Internet of things
Energy meter
Random forest
Homomorphic encryption
