Research On Computing Offloading Scheme Based On Vehicle Group In Vehicular Networks

With the development of mobile communication,wireless sensor and on-board equipment manufacturing technology,the Internet of vehicles has become an important part of intelligent transportation system.The Internet of vehicles has the characteristics of high-speed mobility,fast network topology change,and time-varying wireless channel.With the emergence of a large number of computing intensive applications,the requirements of the vehicular networks for the storage and computing capacity,communication reliability and computation offloading latency of vehicles are becoming higher.In order to deal with the problem of insufficient vehicle computing capacity,researchers propose to combine mobile edge computing(MEC)with the Internet of vehicles,and carry out in-depth research on the key technologies such as computation offloading.Vehicle grouping is a typical scenario of vehicle cooperation,in which vehicles can share information and help the vehicles with insufficient computing resources to offload.However,the vehicle may leave the group in the process of helping others in the middle of completing the computation task,resulting in the lost of returned computation results.On the other hand,when the computation load is heavy in the group,how to select other vehicle groups for computation offloading to meet the latency requirement of computation task is also worth further discussion.In view of the above two problems,this thesis has carried out in-depth research on the computational offloading within and between vehicle groups in the Internet of vehicles.The main work of this thesis is as follows:Aiming at the problem of how to maintain stable offloading in the vehicle group in the vehicular networks,this thesis proposes a computational offloading algorithm which considers the time when the members of the vehicle group can maintain a stable connection and the sudden departure of vehicles.A group of vehicles with the same driving direction can share route information and help to complete the computation task.However,considering that the vehicles in the vehicle group may leave the group in the computation process,it is imperative to consider how to return the computation results.This thesis analyzes the two cases of vehicles leaving the group due to failure and leaving the group on their own initiative.Different types of computing tasks are divided into independent sub tasks,and an collaborative computing task allocation algorithm is proposed.The total latency of computing task offloading includes task upload latency,task computation latency and result return latency.The proposed algorithm achieves the goal of minimizing the total task latency by iteratively adjusting the allocation mode of subtasks.Simulation results show that the algorithm can reduce the overall latency and improve the utilization of vehicle computing resources.Aiming at the problem of heavy computing load in some vehicle groups in the multi vehicle group scenario of vehicular networks,a cooperative offloading strategy between vehicle groups based on MEC is proposed.Due to the different track and speed of different vehicle groups,it is difficult to maintain stable offloading between vehicle groups.The driving conditions of different vehicle groups are considered in modeling.According to the moving models of different vehicle groups,the time period for each vehicle group to keep connected with other vehicle groups is predicted.Each group maintains the connection map and updates it continuously.The task offloading problem is described as a mixed integer programming problem,and the optimization objective is to minimize the average completion latency of all vehicle groups.Considering that the problem is NP-Hard,it is difficult to obtain the optimal solution in polynomial time,so genetic algorithm is used to obtain the suboptimal solution iteratively.Simulation results show that the proposed algorithm can effectively improve the proportion of vehicles involved in offloading and reduce the average task latency of all vehicle groups.