Research On Location Privacy Algorithm Based On Period And Semantics

With the rise of the mobile Internet,devices with positioning functions such as smartphones and PADs have proliferated,and location-based services(LBS)have been used more and more widely,and have gradually changed people’s daily life such as surfing the Internet and traveling,play an extremely important role.The widespread use of LBS has also generated a large amount of location data,and application providers also provide users with more high-quality and personalized services by collecting users’ locations.Massive user location data are stored in cloud servers,and these location data are in Academic fields,commercial fields,social development and other fields have great value,but they also easily pose a huge threat to user privacy because they contain users’ personal information.Existing research directions focus on the spatial attributes of location data,and seldom consider other background information.Some research schemes can only be aimed at specific scenarios,and have relatively large restrictions on application scenarios.Some protect user location privacy across the board,without taking into account the sensitivity of different location data and the differences in privacy needs of different users,resulting in a decline in service quality and making it difficult to meet users’ high service quality requirements.Aiming at these deficiencies,this paper conducts research on privacy and security-related technologies in location-based services.The work and innovations are as follows.First of all,in the existing location privacy protection schemes for real-time single queries,more consideration is given to background knowledge such as the historical query probability of the location and the size of the anonymity set.The historical query probability of the same semantic location point may change dramatically.Considering the query probability of location by time period,semantic information and the size of the anonymity set comprehensively,a fine-grained dummy location generation algorithm TSDLS with different time periods is proposed.The scheme first selects locations with similar query probability to the user’s time period to form a candidate set,then calculates the semantic distance and the formed area between the candidate locations and the dummy location set,and selects the optimal location to join the dummy location set.Experimental results show that the algorithm proposed in this paper can generate effective location privacy protection for users..Secondly,most of the offline user trajectory release schemes process all the trajectory data and add noise to meet differential privacy.The lack of user personalization leads to low data availability and less consideration of the semantics information of location points.Considering the above situations,a personalized semantic-sensitive trajectory publishing algorithm TSDP is designed.This scheme collects the user’s semantic sensitivity when publishing,and counts the frequently visited semantic location points in the user’s trajectory.It also considers the impact on sensitivity due to different time periods,and adds noise that meets differential privacy.Experimental results show that compared with other differential privacy protection methods,this method maximizes the availability of published data while ensuring privacy.Finally,a usable location privacy-preserving system is implemented by combining the time-segment fine-grained dummy location generation algorithm TSDLS with the personalized semantic-sensitive trajectory publishing algorithm TSDP.