Research On Trajectory Data Privacy Protection Approach For Location Based Services

Trajectory data contains abundant geographic and temporal information,and also contains users’ sensitive privacy information.In-depth mining and analysis of trajectory data can support a variety of applications.Among them,Location Based Services(LBS)trajectory data has played an important role in social development.When the mass trajectory data is released directly without processing,although it can provide more convenience,it will also lead to the risk of personal privacy disclosure.Therefore,in the LBS platform,how to ensure the availability of trajectory data without compromising user privacy is crucial.Existing trajectory privacy protection schemes based on different application scenarios of location service still have problems such as low privacy of anonymous sets and poor availability of protected data.Based on this,this paper proposes the trajectory data privacy protection scheme with higher availability for different application scenarios of location services.The main work and achievements are as follows:(1)Aiming at the route search scenario in LBS,existing privacy protection schemes may have the problem of unreasonable construction of trajectory anonymous set,a Privacy Protection Protocol of Trajectory Based on K-anonymity for Route Search is proposed.Firstly,a false trajectory generation algorithm based on distance is proposed to construct an anonymous set of trajectory data combined with distance factor.Secondly,a false trajectory generation algorithm based on distance and angle is proposed to further improve the construction of anonymous set of trajectory data.Finally,a false trajectory generation algorithm based on map matching is proposed to improve the rationality of anonymous set of trajectory data.Experiments show that the proposed protocol can improve service accuracy while ensuring the rationality of anonymous trajectory.(2)Aiming at location recommendation scenario in LBS,existing privacy protection schemes may ignore the influence of semantic location,a Privacy Protection Mechanism of Trajectory Based on Central Differential Privacy for Location Recommendation is proposed.Firstly,a processing algorithm based on semantic location sensitivity is proposed to calculate the semantic sensitivity and privacy level of the location.Secondly,a privacy budget allocation algorithm based on prefix tree and a privacy budget adjustment algorithm based on Markov chain are proposed to achieve a better allocation of privacy budget.Finally,a location recommendation algorithm under differential privacy protection is proposed to verify the availability of the mechanism.Experiments show that the proposed mechanism can improve the performance of location recommendation while protecting data privacy.(3)Aiming at the trajectory publishing scenario in LBS,existing privacy protection schemes may have the problem of poor data availability due to the imbalance of privacy budget allocation,a Privacy Protection Strategy of Trajectory Based on Local Differential Privacy for Trajectory Publishing is proposed.Firstly,a user sampling grouping algorithm is proposed to reduce noise errors.Secondly,a user division algorithm based on three-way decision is proposed to realize automatic user partitioning and reduce the operating cost.Finally,an adaptive privacy budget allocation algorithm based on the water-filling principle is proposed to achieve the optimal allocation of privacy budget.Experiments show that the proposed strategy can improve the availability of published data while ensuring data privacy.In conclusion,this paper realizes the privacy protection of trajectory data for different scenarios in LBS,which improves the availability of trajectory data while ensuring the privacy of trajectory data.