Effective Selection of Targeted Advertisements for Vehicular Users

This paper focuses on targeted advertising for vehicular users, where users receive advertisements (ads) from roadside units and the vehicle onboard system displays only ads that are relevant to the user. A broker broadcasts ads and is paid by advertisers based on the number of vehicles that displayed each ad. The problem we study is the following: given that the broker can broadcast a limited number of ads, what is the strategy for ad selection that maximizes the broker's revenue? We first identify the conflict existing between users' interests and broker's revenue as a critical feature of this scenario, which may dramatically reduce the broker's revenue. Then, given the problem complexity, we propose Volfied, an algorithm that solves this conflict, allows for near-optimal broker's revenue and has very limited computational complexity. Our results show that Volfied increases the broker's revenue by up to 70% with respect to state-of-the-art alternatives.