Delay-Constraint Offloading and Joint Resource Allocation in MEC Based Vehicular Network

Abstract

Cloud-enabled vehicular network is a promising paradigm to meet the ever-increasing demands from various mobile applications. However, huge transmission delay and serious degradation may occur currently due to the long distance away from vehicular terminals. Thus, a new issue named Mobile Edge Computing (MEC) is established as an effective way to provide cloud computing capabilities at the edge of pervasive radio access networks, in close proximity to vehicular terminals. In this paper, we build a MEC based vehicular network framework to solve offloading and resource allocation problem of resource-hungry tasks such as data and computing dual intensive task by collaborative computing in the multi-user single-task environment, in the meantime, reducing latency and improving the resource utilization efficiently. First, we come up with an appropriate task offloading decision for every vehicular terminal to decide the proportion between local computing and mobile edge computing in the form of a delay-constraint objective function by considering both moving pattern of vehicles and channel conditions. Then, a joint resource allocation scheme is designed through solving a two-dimensional knapsack problem. Efficient resource allocation results can be obtained by maximizing the sum of time duration difference between local computing only and collaborative computing. Simulation results are provided to validate the effectiveness of proposed work and demonstrate the impacts of various parameters to the system performance.