Computation Offloading Management for Vehicular Ad Hoc Cloud

摘要

Vehicular Ad Hoc Cloud(VACloud) research tries to combine vehicular ad hoc networks and mobile cloud computing, i.e., to implement cloud computing among vehicles interconnecting with wireless ad hoc networks. This paper proposed a computation offloading framework for VACloud, whose functional components including resource registration and discovery, application partition, surrogate selection, offloading monitoring,etc. One application partition method and four surrogate selection strategies were put forward and their performances were investigated via simulation experiments. It's shown that the multi-attributed strategy that considers multiple attributes, including the computation capacity and the distance, outperformed than the others, with higher task completion rate and shorter completion time.