基于敏感位置多样性的LBS位置隐私保护方法研究

Research of LBS privacy preserving based on sensitive location diversity

针对LBS查询服务中构造的匿名框或选取的锚点仍位于敏感区域而导致的位置隐私泄漏问题,提出了基于敏感位置多样性的锚点选取算法。该算法根据用户访问数量和访问高峰时段,对不同敏感位置进行定义和筛选,选择具有相似特征的其他敏感位置构成多样性区域,并以该区域形心作为查询锚点,提高用户在敏感位置出现的多样性。以该锚点为查询标志,提出一种均衡增量近邻兴趣点查询算法HINN,在无需用户提供真实位置坐标的条件下实现K近邻兴趣点查询,同时改进了Space Twist方法中存在的查询兴趣点围绕锚点分布的缺陷,提高了查询准确度。实验表明,本方法实现了用户在敏感区域停留时的位置隐私保护目标,同时具有良好的兴趣点查询质量和较低的通信开销。

Before getting location-based query service, constructing a cloaking region or picking an anchor which is still in a sensitive area is vulnerable to lead location privacy exposure. An algorithm of selecting anchor is proposed based on sensitive location diversity. By defining sensitive locations and filtering different ones according to users＇ visiting number and peak time, locations with similar features are chosen to construct a diversity zone, and the centroid of the zone is chosen as an anchor that raises location diversity. Referring to SpaceTwist, a query algorithm （H/NN） is proposed to get places of interest （PoI）, and query results can be inferred without providing any user＇s actual location. The defect in SpaceTwist that Pols are found around the anchor is modified, which improves querying accuracy. The experiments show that users＇ location privacy is protected well when the user is staying at a sensitive place, and the method has good work- ing performances.