Research On Trajectory Privacy Protection Technology Based On STPM Model

With the popularization and development of wireless technology,global positioning technology,and mobile facilities,a large amount of trajectory data will be generated when users use location-based services.By mining trajectory data,users’ behavior habits and interests can be analyzed,providing users with better location services.However,users also face various security risks while using location-based services.On the basis of the existing trajectory privacy protection models and algorithms,most of the current trajectory privacy protection schemes only focus on the privacy protection of stay points,do not consider the semantics of the user’s location,and cannot effectively resist semantic attacks.This paper proposes a Trajectory privacy protection method based on the STPM model.First,a new semantic trajectory model is constructed by extracting the spatiotemporal sequences,important spatial points(start and endpoints and stop points),and motion patterns in the location semantics,and the trajectory is fuzzified at the same time.Then,semantic and spatiotemporal similarity comparisons are performed on the obfuscated trajectories to obtain a recordable and releasable set of anonymous trajectories.Finally,through experimental analysis,it is verified that this method can effectively improve the user privacy protection effect.The important work and innovations of this paper mainly include the following aspects:(1)A trajectory privacy protection model STPM is proposed.The trajectory information is extracted from the client and the server for preprocessing.Different from the trajectory privacy protection scheme that only considers the stay point,the information such as longitude,latitude,timestamp,interest point,speed,and motion pattern extracted from the trajectory are used as semantic similarity factors,to build a new trajectory privacy protection model based on location semantics.(2)Construct a sensitive area to blur the trajectory.The sensitive points are processed based on the K-anonymous method,and finally,a sensitive area including K-1 similar sensitive points is formed.In this sensitive area,the trajectory is fuzzified by combining the user’s movement pattern,road weight,and road network structure.Among them,the fuzzy targets mainly include start and endpoints and stop points.The methods of fuzzification can be divided into two types: trajectory segment pruning and trajectory segment addition.(3)Perform similarity comparison to construct an anonymous trajectory set.The temporal and spatial similarity of the trajectory is measured by normalization and multiattribute decision-making,and the similarity between the stop point and the movement pattern on the semantic trajectory is measured by the cosine similarity method.Filter out K-1 trajectories with the highest similarity,and form an anonymous set with the original trajectories.Finally,by publishing an anonymous trajectory set,the performance of user trajectory privacy protection is greatly improved.There are 17 pictures,7 tables,and 69 references.