中文摘要：

基于隐形区的位置混淆技术是实现位置隐私广泛研究的技术，但该技术需要可信第三方且无法防止基于背景信息的推理攻击，容易’泄露位置隐私。针对这一难题，提出了以用户为中心的差分扰动位置隐私保护方法，不需要可信第三方，同时增强了用户位置隐私。该方法采用修改的Hilbert曲线映射技术将地图中用户的每个位置投影到一维空间，通过组合足匿名和差分隐私技术随机产生扰动，并将扰动位置作为用户真实位置提交给服务商。为了解决移动设备资源受限问题，采用基于四分树的方法将用户的上下文存储和转换为比特流，由此获得了有效的时空复杂度和很高的检索准确率。安全分析表明，该方法能有效保护用户位置隐私；实验评估表明，与采用标准Hilbert曲线映射的方法相比检索准确率平均提高了15．4％。所提方法在隐私保护和服务精度之间取得了较好的权衡，对隐私保护系统设计具有一定的理论和实际意义。

英文摘要：

Abstract： A user-centric location privacy-preserving method with differential perturbations （Ulp2mDP） is proposed to solve the problem that the location obfuscation technique using cloaking region requires a trusted third part （TTP） and cannot sufficiently prevent inference attacks based on background information, and hence is easy to leak location privacy. The method can enhance the user＇s location privacy without requiring a TTP. The Ulp2mDP uses a modified Hilbert curve to project each 2-D geographical location of user into a 1-D space, and then randomly generates a reasonable perturbed location by combining the k anonymity with differential privacy technique. The perturbed value is then submitted as the user＇s real location to the service provider. In order to address the resource limitation of mobile devices, a quad-tree based scheme is used to transform and to store the user＇s context information as bit stream,which are highly efficient with respect to time and space complexities, hence to achieves high precision of retrieval. Security analysis shows that the Ulp2mDP can effectively protect user＇s location privacy. Experimental evaluation and a comparison with the approach using standard Hilbert curve show that the average retrieval accuracy of the Ulp2mDP increases by 15.4%. It is concluded that the Ulp2mDP provides a tradeoff between privacy preserving and service accuracy, and has a certain theoretical and practical significance for the design of privacy-preserving systems.