Utility maximization data scheduling in drone-assisted vehicular networks

In Vehicular Networks (VANETs), the rapid movement of vehicles leads to highly time-varying network topology and brings the problem of blind spots in signal coverage. To address these issues, drones are employed to assist data dissemination in VANETs attribute to the flexible development of drones. As random data transmission will result in decreased network performance and low efficiency in data retrieval service, it is urgent to develop efficient data scheduling schemes that meet the quality of service (QoS) of different applications in VANETs. In this context, to fulfill the request of real-time and reliable data transmission, we formulate a data scheduling problem that considers the factors such as priority of data transmission, link quality, link connection time and network fairness. Our goal is to reduce random data transmission, therefore maximizing network transmission utility. We utilize graph theory to describe network topology, in which vehicles and drones are represented as vertices and links between the nodes are represented as edges. The weight of edges indicates the utility obtained when data transmits between the corresponding nodes. We reduce the data scheduling problem to the maximum weighted matching problem and then propose a data scheduling scheme that can satisfy data requests of different vehicles to the maximum extent. The theoretical analysis derives the scheduling algorithm's time complexity and the number of scheduling stages required to satisfy the data requests. Finally, simulation verifies the proposed scheme's effectiveness in terms of service rate, service delay, fairness and throughput.