Location Privacy Protection in Mobile Networks (SpringerBriefs in Computer Science)
Format: PDF / Kindle (mobi) / ePub
This SpringerBrief analyzes the potential privacy threats in wireless and mobile network environments, and reviews some existing works. It proposes multiple privacy preserving techniques against several types of privacy threats that are targeting users in a mobile network environment. Depending on the network architecture, different approaches can be adopted. The first proposed approach considers a three-party system architecture where there is a trusted central authority that can be used to protect users? privacy. The second approach considers a totally distributed environment where users perform privacy protection by themselves. Finally, more general system architecture is discussed including how a semi-trusted server may exist, but users need to collaborate to achieve maximized privacy protection. This brief is designed for researchers and professionals working with privacy preservation, mobile networks, and threat models. The variety of approaches presented makes it useful for students as well.
As long as the user stays within the current service area, these dummy user’s locations will keep updated to LBS server 38 3 Privacy Preservation Using Game-Theoretic Approach Fig. 3.3 A snapshot showing how dummy users are used to protect against the inference attack. Without dummy user u 2 , A may discover the correlation between u 2 and r2 with high probability. With the newly introduced u 2 , the risk of r2 ’s whole trajectory being revealed is reduced as if a real user is traveling.
paid. 3.6 Timing-Aware Dummy User Generation Game 43 3.6 Timing-Aware Dummy User Generation Game In this section, we extend the previously developed DUG game model to incorporate the timing of decisions. Specifically, instead of requiring LBS users that choose Cooperate to generate dummy users as soon as they start traveling in the service area, we consider the case where players may intentionally delay their dummy generation and expect others to generate dummy users before they do. Since
the number of players is higher than 4, the favorable outcome rate of our algorithm is almost 100 %. This justifies our theoretical analysis in Theorem 3.3. When the game reaches the symmetric mixed strategy equilibrium, and the number of players becomes large, it becomes almost impossible for all players to choose the Defect strategy. For the T-DUG game, we further investigate the achieved DoPs as well as the earliest time for dummy user generation. The results are plotted in Fig. 3.6. Comparing
sensor node checks the mobile sink’s ID in the message to determine if it is necessary to create a new trail reference. The procedure is summarized in Algorithm 8. In SinkTrail trail references of each node represent node locations in different logical coordinate spaces, when it comes to data forwarding, because reporting to any mobile sink is valid, the node can choose the neighbor closest to a mobile sink in any coordinate space. Sink location in each logical coordinate space is still [2, 1,
listed to ease the presentation in the later sections (Table 2.1). 2.3.2 Mix Zone Implementation Although the theoretical mix zone model discussed in Sect. 2.1 seems to be effective in protecting a user’s privacy, these two conditions, i.e., k users exist in a mix zone at some point of time and users have random moving paths, may not be easily satisfied in real world, especially on a road network . A significant amount of research [4, 8, 17] has been devoted to investigating the optimal size