RSU-aided Mobility-aware Dynamic Resource Allocation for Vehicular Cloud Services

Efficient allocation of resources among the competing applications can pose many challenges in realizing vehicular cloud services (VCS). First, high mobility of vehicles and frequent change in topology makes formation and maintenance of vehicular cloud difficult. Second, applications representing vehicular cloud services are highly heterogeneous in nature in terms of resource requirements. Finally, available amount of resources at a particular time may be less than the required resources. In this paper, we have studied the behaviour of dynamic resource allocation strategies in a vehicular cloud scenario, where a road side unit (RSU) will coordinate the information gathering and resource allocation among the competing services. We have used bandwidth as a resource and vehicles willing to share excess bandwidth will act as service vehicles. Vehicles requiring bandwidth will act as client vehicles and will broadcast their needs to the RSU mentioning minimum and maximum requirements of bandwidth. The RSU will use dynamic resource allocation algorithms to distribute the client vehicle service tasks to suitable service vehicles. Three algorithms have been proposed based on different performance parameters. We have simulated and tested the performance of the proposed algorithms in different vehicular cloud scenarios and discussed the results to gain some insights on dynamic resource provisioning for vehicular cloud services.