通信学报 ›› 2019, Vol. 40 ›› Issue (5): 88-97.doi: 10.11959/j.issn.1000-436x.2019095

• 学术论文 • 上一篇    下一篇

基于边介数模型的差分隐私保护方案

黄海平1,2,王凯1,2,汤雄1,2,张东军1,2   

  1. 1 南京邮电大学计算机学院,江苏 南京 210023
    2 江苏省无线传感网高技术研究重点实验室,江苏 南京 210023
  • 修回日期:2019-03-21 出版日期:2019-05-25 发布日期:2019-05-30
  • 作者简介:黄海平(1981- ),男,福建三明人,博士,南京邮电大学计算机学院教授、副院长,主要研究方向为物联网安全和数据隐私保护等。|王凯(1994- ),男,江苏扬州人,南京邮电大学硕士生,主要研究方向为物联网安全和数据隐私保护。|汤雄(1992- ),男,江苏沭阳人,南京邮电大学硕士生,主要研究方向为物联网安全和数据隐私保护。|张东军(1993- ),男,江苏徐州人,南京邮电大学硕士生,主要研究方向为物联网安全和数据隐私保护。
  • 基金资助:
    国家自然科学基金资助项目(61672297);江苏省重点研发计划(社会发展)基金资助项目(BE2017742);江苏省六大人才高峰基金资助项目(DZXX-017)

Differential privacy protection scheme based on edge betweenness model

Haiping HUANG1,2,Kai WANG1,2,Xiong TANG1,2,Dongjun ZHANG1,2   

  1. 1 College of Computer,Nanjing University of Posts and Telecommunications,Nanjing 210023,China
    2 High Technology Research Key Laboratory of Wireless Sensor Network of Jiangsu Province,Nanjing 210023,China
  • Revised:2019-03-21 Online:2019-05-25 Published:2019-05-30
  • Supported by:
    The National Natural Science Foundation of China(61672297);The Key Research and Development Program of Jiangsu Province (Social Development Program)(BE2017742);The Sixth Talent Peaks Project of Jiangsu Province in China(DZXX-017)

摘要:

随着社交网络应用的不断发展,用户社交关系等个人隐私数据的安全保护问题亟待解决。为显著减小社交网络数据的敏感度,提出了一种基于边介数模型的差分隐私保护方案 BCPA。基于 dK 模型捕获图结构对应的2K 序列,根据边中介中心性系数对 2K 序列重新排序;依据排序结果将 2K 序列聚类成多个子序列,再利用 dK扰动算法对各子序列分别进行加噪;根据整合后的新 2K 序列生成满足差分隐私的社交网络发布图。基于真实数据集,通过模拟仿真将所提方案与其他经典方案进行比较,实验结果表明,所提方案在保证较强隐私保护性的同时,提高了发布数据的准确性和可用性。

关键词: 社交网络, 隐私保护, 差分隐私, dK模型, 聚类, 分组扰动

Abstract:

With the continuous development of social network application,user’s personal social data is so sensitive that the problem of privacy protection needs to be solved urgently.In order to reduce the network data sensitivity,a differential privacy protection scheme BCPA based on edge betweenness model was proposed.The 2K sequence corresponding to the graph structure based on the dK model was captured,and 2K sequences based on the edge betweenness centrality were reordered.According to the result of reordering,the 2K sequence was grouped into several sub-sequences,and each sub-sequence was respectively added with noise by a dK perturbation algorithm.Finally,a social network graph satisfying differential privacy was generated according to the new 2K sequences after integration.Based on the real datasets,the scheme was compared with the classical schemes through simulation experiments.The results demonstrate that it improves the accuracy and usability of data while ensuring desired privacy protection level.

Key words: social network, privacy protection, differential privacy, dK model, clustering, group perturbation

中图分类号: 

[1] 刘伯涛. 移动回传的融合之路[J]. 电信科学, 2009, 25(11): 91 -93 .
[2] 鲜永菊,董灿,张祖凡,吴东伟. LTE-A载波聚合下的载波切换分析[J]. 电信科学, 2009, 25(12): 46 -50 .
[3] 桑俊俊,石胜飞,李建中,熊蜀光. 无线传感器网络分布式单向链路检测算法[J]. 通信学报, 2008, 29(11): 22 -172 .
[4] 曾 益,胡 波,冯 辉. 用于传感器网络的高效分时洪泛时钟同步协议[J]. 通信学报, 2007, 28(5): 2 -14 .
[5] 王俊波,陈 明. 单业务TDD-CDMA系统上行用户容量分析[J]. 通信学报, 2007, 28(6): 8 -53 .
[6] 牛德华,马建峰,马卓,李辰楠,王蕾. 基于属性的安全增强云存储访问控制方案[J]. 通信学报, 2013, 34(Z1): 37 -284 .
[7] 刘 龙,宋琦军,赵太飞,元向辉. 基于运动矢量时-空特性的快速运动估计算法研究[J]. 通信学报, 2013, 34(1): 14 -127 .
[8] 王亚石,闵丽娟,周严. OSS/BSS一体化及其与ITSM的融合[J]. 电信科学, 2014, 30(6): 17 -23 .
[9] 彭俊宇,蔡孙增,朱正航,徐景,周婷. 基于MIMO-OFDM的高频段Gbit/s通信系统设计和实现[J]. 电信科学, 2014, 30(6): 95 -101 .
[10] 杨春刚,盛敏,董延杰,李建东,李红艳,刘勤. 认知网络中基于网络辅助的速率控制方法[J]. 通信学报, 2013, 34(5): 15 -135 .