高级检索

    基于博弈分析的车辆感知网络节点轨迹隐私保护机制

    Privacy Preserving for Node Trajectory in VSN: A Game-Theoretic Analysis Based Approach

    • 摘要: 利用车辆移动“众包模式”为驾驶员提供实时路况信息和探索新的商业服务模式,是车载感知网络的新型应用之一.然而,车辆移动轨迹中的敏感信息易与用户身份相关联,存在隐私泄露问题.提出了一种车载感知网络节点轨迹隐私保护算法,克服了匿名和隐藏技术缺乏数据真实性权衡的缺点,并考虑到多种攻击策略的影响.对于给定的一系列轨迹集,首先确定边信息概率分布,建立攻击者和防御者模型,通过求解攻防博弈中的纳什均衡选择最优的防御策略,并证明其有效性.在此基础上,折中轨迹数据真实性和隐私性,以最大化防御者效用为目标,提出轨迹隐私保护算法.实验结果表明,防御策略在不同的攻击策略下展现出不同的性能,算法所选择的防御策略优于其他的策略.

       

      Abstract: By using crowdsourcing, vehicular sensor networks (VSN) are considered essential for achieving automatic and dynamic traffic information collection, and have created various fresh new business applications and services in our daily lives. However, the published trajectories that collected in VSNs raise significant privacy concerns. These existing methods, such as anonymization and cloaking techniques, though they are attractive for providing strong privacy guarantees, generally fail to satisfy the accuracy requirements of the trajectory data based applications. In addition, different attack strategies will result in quite different performance under various privacy preserving strategies. In order to address these challenges, we present a location privacy protection method, the DefenseGame algorithm. Given a set of trajectories and a probability density function for side information, the algorithm can assist the defender in selecting the optimal defense strategies by calculating the equilibriums in attack and defense games. In the attack and defense game, we use a game-theoretic model to capture the behavior of the adversary and defender, and we show the effectiveness of the two kinds of defense strategies in the adversary’s inference attacks. Our experimental results indicate that the same defense strategy shows different performance for attack strategies and the proposed algorithm can help to obtain higher defender’s utility compared with other approaches.

       

    /

    返回文章
    返回