Research On User's Location Privacy Protection Algorithm For Road Networks

With the rapid development of wireless communication technology, as the representative of mobile terminal devices, smartphone has become extremely common, People use these devices to obtain information conveniently through location based services. But at the same time of enjoying LBS, the disclosure of location privacy has become a very serious problem. If the attacker obtains the accurate location of user, the other personal privacy related to position will be exposed. So, it is necessary to introduce location privacy protection mechanism. Currently, researchers have proposed many location privacy protection algorithms. However, the majority of them assume that the mobile users move in free space and ignore the objective law that users move along road networks. Based on the attack situation existing in road networks, the reaseach of location privacy protection algorithm is carried out in this thesis.Based on the adequate analysis of the common road networks system architecture's advantages and disadvantages, this thesis proposes a central server distributed deployment method. Trusted authority(TA) selects the deployment location for server based on three strategies, and divides the whole road networks by the deployment location. The Voronoi map is beneficinal to anonymous work.Aimming at snapshot LBS in road networks, this thesis presents a V-K location privacy protection algorithm. The algorithm adopts the way of collective k-anonymity, and satisfies all the users' customized anonymity requirements. To deal with edge-weighed attack, V-K algorithm sort the segments for each Voronoi zone in ascending order based on the weight of segment, and construct candidate set by selecting the users on the neighboring segment and the other users on current segment. To deal with replay attack, V-K algorithm randomly selects user from candidate set, and adds it in anonymity set, so as to reduce the probability of replay attack. In order to reduce the anonymity set's space span, V-K algorithm judges the spatial tolerance before adding the user in the anonymity set. Simulation experiments show that the V-W algorithm can effectively resist attacks and also has advantages in terms of query cost.Aimming at continuous LBS in road networks, this thesis presents a continuous LBS location privacy protection algorithm based on Voronoi zone prediction. To deal with the possible location privacy exposure caused by directly transplanting snapshot LBS into continuous LBS, this thesis predicts the next Voronoi zone for the registered user. This thesis still adopts collective anonymity in the algorithm design. To deal with replay attack, this algorithm randomly selects user from Voronoi zone, and adds it in anonymity set. To deal with continuous query attack, this algorithm selects the user whose prediction result for next Voronoi zone is the same as the query user, and adds it in anonymity set. In order to improve the success rate of anonymity as much as possible, this algorithm sorts the neighboring V zone in ascending order by the distance between query user and the center of neighboring V zone, and mergers the Voronoi zone whose value is low.