Online Resource Allocation in Internet of Vehicles Using Topology Attribute-Aware Genetic Algorithm

Virtual resource allocation, widely referred to as Virtual Network Embedding (VNE), has received increasing attention from both industry and academia. In fact, VNE has ubiquitously become a technological leap in Internet of Vehicles (IoV) which is a fundamental framework for the anticipated success of future intelligent transportation. The general VNE problem has been shown to be NP-hard [1], [2] and finding an optimal VNE in a dynamic environment like IoV is even more challenging. In fact, research on VNE in dynamic environments, where connected moving vehicles act as substrate nodes to provision requested services, is still in its nascent stages. As a result, this paper proposes a Genetic Algorithm (GA) assisted by a novel fitness function considering critical network topological attributes and resource constraints for dealing with the online VNE problem considering vehicle mobility. Simulation results based on the Random Waypoint (RWP) mobility model indicate that the proposed algorithm achieves better performance compared to several existing VNE algorithms.