Graph-based optimal revenue packet scheduling in Vehicle-to-Infrastructure communication

Recently, LTE-based Vehicle-to-Infrastructure (V2I) communication is widely studied due to its considerable potential to satisfy users' Quality of Service (QoS) requirements. It is convenient within the diversity of infrastructures. In particular, packet scheduling is of great importance in V2I. In this paper, we establish an optimization packet scheduling model in V2I according to the user's requests including lifetime, the number of needed packets and individual costs. Then, it is designed where the cost of each packet obeys linear decreasing function in a given time interval in order to be closer to the reality. Furthermore, we divide requirements with respect to the number of desired data packets to transform the packet delivery problem to a maximum weight problem in bipartite graph. At the same time, Kuhn-Munkres (KM) algorithm is adopted to maximize the revenue while reducing the complexity. The simulation results show that our proposed algorithm is effective both in offline and online case which increases 52.31% overall revenue while reducing 46.80% CPU time.