An Energy-Aware Service Management Algorithm for Vehicular Cloud Computing

Abstract

AbstractThe rapid technological evolution of cloud computing and technical innovation in the intelligent transportation system have allowed vehicular cloud computing (VCC) to gain massive popularity across industry and academia. Recent studies have shown that modern vehicles have the potential for hosting virtual machines (VM) to augment the services of cloud service providers (CSP). Nowadays, vehicles are powered by alternative fuels (i.e., green vehicles (GVs)) that generate less carbon footprint than conventional vehicles (CVs). The cost of executing a User Request (UR) by a VM hosted on a GV is less than a CV. However, GVs are subjected to constraints, such as irregularity in energy availability, limited processing capability and lower bandwidth capacity. The submitted URs by the cloud users have different quality of service (QoS) requirements. QoS defines the priority, promised completion time and promised cost between the CSP and the cloud users. In this paper, we propose energy-aware service management (EASM) algorithm that considers different types of energy sources of vehicles and priority of the URs, to map and migrate them to the vehicles on the move, vehicles in the parking lot and roadways. We compare the EASM algorithm with the existing vehicular VM migration-uniform (VVMM-U), and mobility and destination workload aware migration (MDWLAM) algorithms through an extensive simulation process. Simulation results indicate that the proposed algorithm outperforms the existing algorithms in terms of four performance metrics, such as total cost, cost of completed URs, profit of completed URs, and percentage of completed high priority URs.KeywordsEnergy-aware service managementVehicular cloud computingRoadside unitGreen vehicleVirtual machineUser request