Trajectory Privacy Preserving Methods For Indoor Moving Objects

With the rapid development of indoor positioning technology,indoor location based services will become a hot industry in the future years,and it will gradually appear in every corner of daily life.And the privacy disclosure caused by indoor location based services is an noteworthy problem.However,research on trajectory privacy preserving for indoor moving objects can solve this matter effectively,and it has a positive effect on development and promotion of indoor location based services.This thesis studies the trajectory privacy preserving methods for indoor moving objects,which based on the existing researches of outdoor trajectory privacy preserving and combined with the indoor trajectory frequent patterns.It is necessary to transform digital trajectory into symbolic trajectory by grid partition in the couse of mining trajectory frequent patterns.In view of the boundary problem in grid partition,this thesis proposes a trajectory frequent patterns mining algorithm based on vague grid sequences(VGSPrefixSpan),which solves the boundary problem by dividing the grid into explicit zone and vague zone based on vertical projection distance.And it mines the trajectory frequent patterns after converting the trajectories to vague grid sequences.Experimental results show that VGS-PrefixSpan algorithm can obtain a better result and has a higher mining efficiency than VSP-PrefixSpan algorithm at the same area ratio.Considering the diverse and complex indoor constraints of indoor environment as well as the attackers with background knowledge,this thesis modifies the calculating formulas of short term disclosure risk and long term disclosure risk.And a dummy trajectory method based on trajectory frequent patterns(TFP-DT)is proposed to protect the trajectory privacy of indoor moving objects.TFPDT algorithm generates dummy trajectories by combining the trajectory frequent patterns and history trajectories,which can reduce the risk of the dummy trajectories be identified.Experimental results show that TFP-DT algorithm can protect the trajectory privacy of indoor moving objects better,and the dummy trajectories generated by TFP-DT algorithm are satisfy indoor constraints more effectively.For the disadvantages of a single category of trajectory privacy preserving algorithm,a hybrid trajectory privacy preserving algorithm(HTPPA)proposed in this thesis,it combines the generalized method and dummy data method.By constructing static anonymous areas and using an anchor point as the query position,it can solve the problem of indoor constraints effectively.While constructing the anonymous user set,some dummy users are added to improve the effect of privacy preserving and users' service experience.Experimental results show that HTPPA can provide users with better effect of trajectory privacy preserving to indoor moving objects at the expense of a small pressure of intermediate server.