Optimizing Task Offloading and Resource Allocation in Vehicular Edge Computing Based on Heterogeneous Cellular Networks
Xinggang Fan, GU Wen-ting, Changqing Long, Chaojie Gu, Shibo He
Abstract
5G is a promising technology for improving the Quality of Service (QoS) in Internet of Vehicles (IoV) applications, including Vehicular Edge Computing (VEC). However, 5G networks have a limited communication range due to their radio-frequency properties, which can be a challenge in dynamic IoV environments. To address this issue, we propose a VEC architecture based on heterogeneous cellular networks, in which vehicles can select the appropriate communication network by classifying tasks according to their maximum tolerable latency. In order to further enhance the overall performance of the VEC system, we developed an efficient scheme that optimizes task offloading decisions and computation resource allocation in the proposed architecture. We analyze and formulate the optimization problem and use the linear relaxation improved branch-and-bound algorithm to solve it. Through extensive simulations, we demonstrate that the proposed scheme is superior to other solutions in computing latency, energy consumption, and failure rate.