Research On Dual Task Offloading Strategy In Vehicle Edge Computing With Parking Assistance

In recent years,Internet of Vehicles(Io V)has become one of the hot topics in academia and industry.With the rapid development of the Internet of vehicles,more and more intelligent vehicles are appearing on the road and generating many computationally intensive applications,such as autonomous driving,real-time online gaming and virtual reality.However,the vehicles themselves have limited computing and storage resources to handle these computationally intensive applications in a timely manner.Cloud Computing(CC)servers have strong Computing and storage capabilities.It is an optional solution to unload computance-intensive application tasks to Cloud servers.However,since CC servers are usually deployed far away from vehicles,the data transfer process consumes a lot of time and may not meet the requirements of real-time task processing.Into this was born Vehicular Edge Computing(VEC).VEC is based on Mobile Edge Computing(MEC),which is close to Mobile end users and pushes Computing power from cloud servers to the Edge of the network near wireless access networks.VEC can offload computation-intensive applications from local vehicles to resource-rich edge servers for execution,speeding up data processing times.However,the number of edge servers deployed and the resources of edge servers are limited.At the same time,although smart vehicles are equipped with high-performance computing units,they are underutilized.In this case,how to design a task unloading strategy in vehicle edge computing to make full use of the idle resources of intelligent vehicles,make up for the limited edge server resources,and reduce the total task completion time and energy consumption has become an urgent problem to be solved.On this basis,this paper studies the dual-task unloading strategy of vehicle edge calculation assisted by parking.The main work is as follows:Firstly,this paper proposes to organize roadside parked vehicles into parking clusters as virtual edge servers to assist the actual deployed edge servers to perform compute-intensive tasks generated by moving vehicles.Virtual edge server makes reasonable use of stable and sufficient computing resources of intelligent vehicles parked on the roadside and effectively solves the problem of limited edge server resources in vehicle edge computing.Second,this paper designs a top-down task scheduling strategy to ensure that edge server can complete the unloading task on time.Moving vehicles on the road generate several delay-sensitive tasks that need unloading calculation,and the task unloading request is uniformly sent to the base station,which performs the proposed top-down task scheduling algorithm.Finally,by solving the two sub-problems of multidimensional multiple-choice knapsack and 0-1integer programming,computing resources and edge server nodes are allocated to all unloading tasks.This assignment strategy ensures the optimality of global task assignment,improves the task completion rate and reduces the task completion time.Thirdly,based on the top-down task scheduling strategy,this paper maximizes the income of parked vehicles through the internal task allocation of parking cluster for the tasks assigned to unload the parking cluster.Based on the intrinsic reward,time penalty function and energy consumption of parked vehicles,the benefits of parked vehicles were expressed as utility values and calculated by utility function formula.The intra-cluster task allocation reduces the energy cost of unloading tasks and ensures the minimum delay of tasks.A large number of simulation results show that the performance of the dual task unloading scheme is significantly higher than that of other comparative strategies.In addition to reducing the task unloading completion time,the revenue and energy consumption of parked vehicles are considered,which improves the experience quality of on-board users.