具有隐私保护功能的交通运输信息管理系统的研究与实现

发布时间：2018-04-13 21:45

本文选题：高速路信息 + 隐私保护　；参考：《北京邮电大学》2016年硕士论文

【摘要】：高速公路作为现代化城市的基础设施,对人们的日常生活和国家经济有着重要的影响。过去“重建设、轻养护”的传统观念,已不适应我国现有高速公路的发展趋势。随着信息技术应用的不断普及和深入,利用信息化手段高效管理高速公路便具有了十分重要的意义。在信息碎片呈井喷式发展的今天,防止隐私泄露、保护个人信息安全便成为衡量应用系统的一个重要指标。本文将差分隐私保护方法与高速公路信息管理相结合,针对常用差分隐私保护框架PINQ中存在的隐私泄露问题提出改进算法,并开发出针对高速公路信息的交通运输信息管理系统。PINQ框架是差分隐私保护方法的一种常用实现框架,其中Apply算法是核心关键算法,但现有Apply算法并没有对查询次数进行限制,存在攻击者可以对大量查询结果进行统计从而得到隐私信息分布区间的可能性,造成隐私泄露。本文对其进行改进,并在此基础上设计实现了增强隐私保护的交通运输信息系统。论文通过实验得到关键参数查询预算的最佳组合值,并且通过分别将改进前后的Apply算法应用于本系统比对来验证改进后的算法具有更强的隐私保护功能;最后将本文提出改进的算法和传统K-匿名算法分别应用于交通运输信息系统中,通过误差分析、隐私泄露情况以及数据可用性等方面的对比,证明本文提出改进的差分隐私保护算法更优。
[Abstract]:Expressway, as the infrastructure of modern city, has an important influence on people's daily life and national economy.In the past, the traditional concept of "paying more attention to construction than maintenance" has not adapted to the development trend of China's existing highways.With the popularization and deepening of the application of information technology, it is of great significance to use information technology to manage expressway efficiently.Nowadays, with the development of information fragments, preventing privacy leakage and protecting personal information security have become an important index to measure the application system.In this paper, the differential privacy protection method is combined with highway information management, and an improved algorithm is proposed to solve the privacy leakage problem in the commonly used differential privacy protection framework (PINQ).This paper also develops the transportation information management system .PINQ framework for highway information, which is a common implementation framework of differential privacy protection method. Apply algorithm is the key algorithm, but the existing Apply algorithm does not limit the number of queries.There is the possibility that an attacker can calculate a large number of query results to obtain the distribution of privacy information, resulting in privacy disclosure.On the basis of the improvement, this paper designs and implements a transportation information system to enhance privacy protection.The best combination value of the key parameter query budget is obtained by experiments, and the improved Apply algorithm is applied to the system to verify that the improved algorithm has stronger privacy protection function.Finally, the improved algorithm and the traditional K- anonymous algorithm are applied to the transportation information system respectively, and the comparison of error analysis, privacy disclosure and data availability, etc.It is proved that the improved differential privacy protection algorithm is better in this paper.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：U495;TP309

【参考文献】