Research On LBS Privacy Protection Based On K-anonymity

Continued advances in mobile networks and positioning technologies have created a strong market push for location-based applications.How to use the location-based service without reveal too much location information has gradually become a hotspot research on location-based service.K-anonymity is an effective means to solve the issue of privacy disclosure in the location-based service,but the research on the personalized K-anonymity mainly allowing the user to choose different privacy levels based on different requirements.However,the user cannot achieve the optimal balance between privacy level and service quality on his own knowledge as the decision is too complicated to work out.In addition,the server which in the location-based service needs to obtain the exact location information to improve the accuracy of the query.Once the server is attacked,the location data will be revealed,thereby digging out the user's private information and harming the interests of user.In order to address these problems,this thesis proposes a personalized K-anonymity recommendation model to help users take the initiative in the trade-off between privacy level and service quality,and proposes a query processing model that does not reveal user's location information when he use the location-based service.Specific work as follows:(1)Analyze the factors that affect the user's decision when he choose the privacy level.(2)Propose a K-anonymity commendation model to help user set the optimal privacy level.Firstly,get a K value based on the user's past privacy preferences.Secondly,considering the factors that affect the user's privacy level,get an ideal K value through continuous feedback.Finally recommend the ideal K value to the user,and the system does not interfere with the user's final choice.(3)Analyze how the attacker obtains the location data of the user,When the attacker obtains the location data from the location-based database,propose a query processing model that does not affect user's privacy information.(4)The query algorithm is given based on the query processing model,and the correctness of the algorithm is verified.The experimental results show that the proposed query processing model can handle the private queries which cannot be completed by the traditional server,and the model can also reduce the query time through the filter to achieve better performance.