期刊
IEEE INTERNET OF THINGS JOURNAL
卷 6, 期 2, 页码 1530-1540出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JIOT.2018.2842773
关键词
Data clustering; differential privacy; Internet of Things (IoT); k-means algorithm; privacy protection
资金
- National Natural Science Foundation of China [61502489, 61402109, 61370078, 61502102, 61502103]
- Chongqing Key Laboratory of Optical Communication and Networks Research Fund [KLOCN2018001]
The ever-growing demand for electrical energy of sensing devices in the Internet of Things (IoT) has led to generating large amounts of electricity consumption data. Electricity service providers often use wireless sensor networks to collect sensing devices' electricity consumption data for statistical analysis, so as to provide sensing devices with improved electrical services. As an important data mining technique, while data clustering excels in dealing with such massive data, it imposes the risk of privacy disclosure in the process of data clustering. In an effort of solving this problem, Blum et al. proposed a differential privacy k-means algorithm, effectively preventing privacy disclosure. However, the availability of data clustering results is reduced due to the data distortion in Blum's algorithm. In this paper, we propose a privacy and availability data clustering (PADC) scheme based on k-means algorithm and differential privacy, which enhances the selection of the initial center points and the distance calculation method from other points to center point. Moreover, PADC attempts to reduce the outlier effect through detecting outliers during the clustering process. Security analysis indicates that our scheme satisfies the goal of differential privacy and prevents privacy information disclosure. Meanwhile, performance evaluation shows that our scheme, at the same privacy level, improves the availability of clustering results compared to the existing differential privacy k-means algorithms, suggesting that our proposed PADC scheme outperforms others for intelligent electrical service in IoT.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据