Research On Location Privacy Protection For Location-based Services

The ubiquitous adoption of smartphones embedded with positioning systems,which are associated with the accessibility of digital maps,makes the location-based services(LBSs)spring up enormously as mushrooming.LBSs has become a new service model.In the past few years,we have witnessed the widespread proliferation of LBSs.Many related products about LBSs have come into being.People access the wealth of personalized services offered by LBSs by downloading associated apps from software platforms,including navigating,socializing,gaming and more.Despite the unprecedented convenience and promising market scale of LBSs,this also triggers elevated privacy concerns.For example,an adversary could infer privacy-sensitive information based on locations users have visited,such as where users work,live,and socialize.Therefore,this paper proposes two location privacy protection methods for the current location privacy leakage problem of LBSs.The main contribution of this paper includes the following two-fold:1.A location privacy protection method based on sensitive locations is proposed.To address the problems that location anonymizer-based privacy mechanisms in smart healthcare systems are vulnerable to performance bottlenecks and subjected to essentially compromise,this paper proposes an adaptive dummy-based mechanism(ADBM)to protect patients' sensitive location information.In this mechanism,patients need to pre-define privacy-sensitive locations that they do not want to disclose to anyone.We use a superGaussian two-dimensional function to determine the perturbation location,which is centered on the predefined sensitive location.The vertical distance between the generated perturbation trajectory and the sensitive location is also considered,ensuring that the sensitive locations are kept at a safe distance from the perturbation trajectory.Our proposed ADBM scheme achieve location privacy while maintaining acceptable accuracy of location information.2.A location privacy protection method based on query probability is proposed.An adversary could infer the true location based on the query probability of the location,inferring the true location from the correlation between consecutive locations in continuous queries of LBSs.Aiming at this challenge,we propose a novel dual privacy protection scheme(DPPS)to enhance privacy while resisting collusion and inference attacks.The DPPS scheme consists of two aspects.First,a hidden Markov model is leveraged for simulating the correlation between consecutive locations in accordance with user mobility.We exploit such Markov process to enhance the user location privacy.In addition,kanonymity is utilized to construct cloaking region to provide query probability anonymity.We generated realistic and indistinguishable dummy locations that are difficult for the adversary to distinguish in cloaking region.A numerical analysis and an empirical evaluation are further performed to validate the effectiveness and efficiency of our DPPS scheme.