Computing Offloading in Vehicular Edge Computing Networks: Full or Partial Offloading?
Chunli Ren, Guoan Zhang, Xiaohui Gu, Yue Li
Abstract
With the surge of vehicular users and the development of mobile applications, the management of big data has become a challenge for the Internet of Vehicle (IoV). Fortunately, mobile edge computing (MEC) bringing the cloud computing and storage resources to the edge of mobile access network is a potential solution, which is capable of effectively reducing task response time and alleviating the pressure of the backhaul network. This paper proposes a joint computing offloading and transmit power allocation scheme for vehicular edge computing networks. Specifically, the vehicle with limited computing capacity and battery lifetime offloads part of computing tasks to MEC server deployed the roadside unit for processing. The offloading data size and the transmit power are jointly optimized to minimize the energy consumption of the vehicle, under the constraint of task response time. Through the analysis of the relationship between the energy consumption and latency of completing the task, we obtained the channel gain threshold for vehicle performing full offloading or partial offloading. Moreover, for the partial offloading, the optimal offloading data size and transmit power and obtained with the convex optimization method. The superiority of the proposed offloading strategy is verified by numerical results, and the effects of some important parameters on the performance of our strategy are also illustrated.