EnTruVe: ENergy and TRUst-aware Virtual Machine allocation in VEhicle fog computing for catering applications in 5G

Abstract

It is undoubted that fog computing contributes in catering the latency-stringent applications of 5G, and one of the enabling technologies that fundamentally ensures the success of fog computing is virtualization as it offers isolation and platform independence. Although the emergence of vehicle-based fog (referred to as v-fog) facilities can certainly benefit from these desirable features of virtualization, there are several challenges that need to be addressed in order to realize the full potential that v-fogs can offer. One of the challenges of virtualization in v-fog is Virtual Machine (VM) migration. There are several factors that trigger a VM migration in a v-fog such as vehicle resource depletion. VM migrations would not only lead to nonessential usage of valuable resources (e.g. energy, bandwidth, memory) in the v-fogs, but also incur various overheads and performance degradation throughout the whole network. Thus, minimizing VM migrations is necessary. Furthermore, to ensure the seamless VM migrations between v-fogs, trust of v-fogs is required. While there exists studies of trust in the virtualization of cloud, they are irrelevant to v-fogs as v-fogs are different in nature (i.e. heterogeneous, mobile) from the cloud. Additionally, trust is not included in the decision making mechanisms of VM allocation for vehicular environments in the existing works. Moreover, as vehicle resources are constrained, their energy has to be utilized efficiently. In this paper, we propose EnTruVe, an ENergy and TRUst-aware VM allocation in VEhicle fog computing solution that aims to minimize the number of VM migrations while reducing VM processing associated energy consumption as much as possible. The VM allocation algorithm in EnTruVe provides a larger selection pool of v-fogs that meets the VMs requirements (e.g. trust, latency), thereby ensuring higher chances of success of VM allocation. Using Analytic Hierarchy Process (AHP), the proposed EnTruVe solution evaluates the v-fogs based on a set of metrics (e.g. energy consumption and end-to-end latency) to select the optimal v-fog for a VM allocation. Results obtained demonstrate that EnTruVe has the least number of VM migrations and it is the most energy efficient solution. Additionally, it shows that EnTruVe provides the highest utilization of v-fogs of up to 57.6% in comparison to other solutions as the number of incoming requests increases.