Privately Information Sharing with Delusive Paths for Data Forwarding in Vehicular Networks

Abstract: We discuss how to efficiently forward data in vehicular networks. Existing solutions do not make full use of trajectory planning of nearby vehicles, or social attributes. The development of onboard navigation system provides drivers some traveling route information. The main novelty of our approach is to envision sharing partial traveling information to the encountered vehicles for better service. Our data forwarding algorithm utilizes this lightweight information under the delusive paths privacy preservation together with the social community structure in vehicular networks. We assume that data transmission is carried by vehicles and road side units (RSUs), while cellular network manages and coordinates relevant global information. The approximate destination set is the set of RSUs that are often passed by the destination vehicle. RSU importance is raised by summing encounter ratios of RSUs in the same connected component. We first define a concept of space-time approachability which is derived from shared partial traveling route and encounter information. It describes the capability of a vehicle to advance messages toward destination. Then, we design a novel data forwarding algorithm, called approachability based algorithm, which combines the space-time approachability with the social community attribute in vehicular networks. We evaluate our approachability based algorithm on data sets from San Francisco Cabspotting and Shanghai Taxi Movement. Results show that the partially shared traveling information plays a positive role in data forwarding in vehicular networks. Approachability based data forwarding algorithm achieves a better performance than existing social based algorithms in vehicular networks.