A Task Assignment Scheme for Parked-Vehicle Assisted Edge Computing in IoV

Vehicular edge computing (VEC) has been envisioned as an important application of edge computing in vehicular networks. Parked vehicles with embedded computation resources could be exploited as a supplement for VEC. They cooperate with edge severs to process offloading tasks at the vehicular network edge, leading to a new paradigm called parked-vehicle assisted edge computing (PVEC) in the Internet of Vehicles (IoV). However, recent researchers mostly focus on how to optimize the total cost of requesting vehicle (RV), and rarely pay attention to the optimization of the utility of PVs that provide services, including the reward from RV and the overhead of executing task. In this paper, we study a task assignment problem with computing delay constraints for PVEC in IoV. Specially, extra performance loss caused by offloading subtasks to PVs is taken into the cost function of RV. The optimal task assignment problem is formulated and solved with the Stackelberg game framework and a ternary search-based algorithm to minimize the cost of RV and maximize the utility of PVs. Finally, extensive numerical results are provided to demonstrate that our scheme is more efficient in deducing the total cost of RV and increasing the reward for PVs than other two existing schemes.