An Adaptive and Dynamic Allocation of Delay-sensitive Vehicular Services in Federated Cloud

Advanced vehicular services demand enormous resources and require low latency. Traditional vehicular networks fail to satisfy such increasing requirements. A new paradigm called Vehicular Cloud Computing (VCC) has been devised to host such services in the cloud. Using the federated cloud concept in this paradigm can result in scalable and cost-efficient solutions with better Quality of Service (QoS). In a federated cloud, many cloud providers share resources to serve a given request. However, resource allocation for such delay-sensitive services in a federated cloud is a challenge. Additionally, vehicle mobility makes the problem challenging as the service provider should monitor the performance of Virtual Machines (VMs) and migrate them when required. In this work, we address the problem of dynamic resource allocation for hosting delay-sensitive vehicular services in a federated cloud. We aim to maximize the number of served requests by meeting their delay requirements while minimizing VM migrations. We introduce a proactive solution that migrates the VMs before violating the actual delay threshold. We propose a delay-aware resource allocation method that considers an adaptive delay warning threshold for various users. Experimental results show that our proposed algorithm fulfills the delay requirements of vehicular services, and it achieves a significant reduction in the number of VM migrations.