While spatial crowdsourcing has become a popular paradigm for spatio-temporal data collection, location privacy has raised increasing concerns among the participants of spatial crowdsourcing projects in recent years. The question of how to implement a spatial crowdsourcing project at minimal cost while preserving location privacy, is the major issue that most existing works have investigated. In this paper, we propose a novel privacy-preserving method for spatial crowdsourcing that combines location obfuscation and path optimization in order to provide enhanced privacy preservation at a minimal cost. We apply geo-indistinguishability and exponential mechanism to achieve an enhanced privacy guarantee. Moreover, because a higher privacy level consistently leads to extra distance cost, we therefore present a path optimization algorithm that reduces the total distance of a spatial crowdsourcing project. The experimental results demonstrate that the proposed method outperforms the traditional methods in terms of privacy level and performance costs.

尽管空间众包已成为时空数据收集的流行范例，但近年来，位置隐私已引起空间众包项目参与者越来越多的关注。如何以最小的成本实施空间众包项目，同时保护位置隐私，是大多数现有作品研究的主要问题。在本文中，我们提出了一种新颖的用于空间众包的隐私保护方法，该方法将位置混淆和路径优化相结合，以便以最小的成本提供增强的隐私保护。我们应用地理不可区分性和指数机制来实现增强的隐私保证。此外，由于较高的隐私级别始终会导致额外的距离费用，因此，我们提出了一种路径优化算法，可以减少空间众包项目的总距离。实验结果表明，该方法在隐私级别和性能成本方面均优于传统方法。