Modeling and Accelerating of Mobile Advertisement Dissemination in Vehicular Networks

Abstract

Mobile advertising is an appealing application in vehicular networks, by which relevant information can be delivered to (and then accessed by) vehicles and pedestrians in different regions. However, due to the complex interplay of mobility and data communication in the dissemination process, it is still not clear how likely an advertisement can be accessed at time t in a particular region, and how to accelerate the dissemination with a limited budget. To solve these problems, we establish a data dissemination model, rigorously define the stage of a dissemination process, and propose a total variance metric to quantitatively differentiate the impact of vehicle mobility from that of V2V communication. Simulations based on two real datasets of taxi traces show that the choice of initial advertisers (i.e., seeds) greatly affects the stage of the dissemination process, and two practical seed selection strategies are proposed as an application of the dissemination model.