Efficient Advertisement Dissemination Strategy With Maximization Influence for Intelligent Networked Vehicle

With the development of computing and communication technology, the popularity of Vehicle-to-Everything (V2X) has promoted the research of advertising among vehicles. The problem of vehicle advertising communication is similar to the problem of Influence Maximization (IM) considered in social networks as both prioritize a set of candidate nodes as the initial seed nodes for the initial information dissemination to maximize the influence of the information in the global network. As user nodes in vehicle networks, vehicles have high mobility, which leads to the highly complex and unstable topology of vehicle networks. Therefore, we propose a practical solution to solve the problem of advertising dissemination among Intelligent Networked Vehicles (INVs). In our strategy, advertisements are first sent to a group of selected seed INVs and then spread among INVs. To encourage INVs to disseminate advertising actively, we propose an incentive scheme that combines online and offline to maximize profit. We propose a dynamic seed INVs replacement algorithm using the Deep Q-Network (DQN) algorithm to decide the update cycle of the seed INVs set. Simulation results show that the proposed strategy is superior to the existing methods in terms of delivery ratio of advertisement dissemination and expected profit of merchant.