Service Recommendation Middleware Based on Location Privacy Protection in VANET

With the help of location-based services (LBS), it makes driving more convience for drivers. However, because the untrusted LBS server may leak the user's location information, the user's privacy is threatened. Moreover, the existing methods of location privacy protection do not take into account the impact of context on privacy protection demand. In addition, heterogeneous data sensed by vehicles also increases the complexity of application development. In order to solve the above problems, we propose a context-based location privacy protection middleware architecture, named PP-OSGi. The middleware simplifies application development by shielding the heterogeneity of vehicle sensed data and upper-layer applications. Furthermore, in order to protect the real location information of service request vehicles under different vehicle densities in a certain area, we propose a dynamically adjustable k-anonymous (DAK) algorithm and a location privacy protection (DLP) algorithm based on a dummy location, which are all encapsulated in PP-OSGi. The DAK and DLP algorithms dynamically determine the location privacy protection strength in different contexts based on the user's location privacy preference model, select an anonymous group of neighboring vehicles to construct a anonymous area, and obtain a dummy location of the service request vehicle. The experimental results show that, under the premise of protecting the location privacy of vehicles, the success rate of service requests is improved and the communication cost between vehicles is reduced.