Providing reliability-aware virtualized network function services for mobile edge computing

Mobile Edge Computing (MEC) has emerged as a promising paradigm to address the conflict between increasing computing-intensive applications and resource-constrained mobile Internet-of-Thing (IoT) devices with portable size and storage. In MEC environments, Virtualized Network Functions (VNFs) are deployed for provisioning network services to users to reduce the service cost on top of dedicated hardware infrastructures. However, VNFs may suffer from failures and malfunctions while network service providers have to guarantee continuously reliable services to their users to meet ever-growing service demands of the users. In this paper, we focus on reliable VNF service provisioning in MECs, by provisioning primary and backup VNF instances in order to meet the reliability requirements of mobile users. We first formulate a novel VNF service reliability problem with the aim to maximize the revenue collected by admitting as many as user requests while meeting individual user service reliability requirements. We then develop two efficient on-line scheduling algorithms for the problem under two different backup schemes: on-site (local) and off-site (remote) schemes, by adopting the primal and dual updating technique. Particularly for the on-site scheme, the proposed on-line algorithm achieves a provable competitive ratio with bounded moderate resource violations. We finally evaluate the proposed algorithms through experimental simulations. The experimental results demonstrate that the proposed algorithms are promising, compared with existing baseline algorithms.