Research On Task Offloading And Resource Allocation Algorithm For MEC-based Internet Of Vehicles

With the emergence of various intelligent vehicle applications that can improve driving experience and traffic safety,on-board equipment can not candle these computing intensive applications with their limited computing resources.The vehicles access to the wireless network through on-board communication equipment in the high-speed mobile environments,and forms the wireless communication links for vehicle-to-vehicle,vehicle-to-infrastructure,and vehicle-to-pedestrian communications.Based on the vehicle network,it is considered as a feasible solution to offload the computing tasks to the cloud server with sufficient computing resources.However,the offloading tasks need to be transmitted through the core network because of the remote deployment of the cloud server,which increases the transmission delay and instability.Multi-access edge computing(MEC)technology provides users with required services and cloud computing functions by deploying MEC platform on the network access side which is closer to users.Compared with cloud computing,it reduces the process of data transmission in the core network,and has the advantages of low latency,high bandwidth and high reliability.In an MEC-based Internet of vehicles system,the offloading decision of users' and the computing resources allocation of MEC platform will influence the task processing efficiency of the whole system.In addition,the allocation of spectrum resources is another focal point.This thesis mainly studies the task offloading strategy and the spectrum and computing resources allocation in MEC-based internet of vehicles system.Firstly,the problem is modeled by network model,communication model and computing model,for the computing tasks generated by on-board applications,three kinds of task processing modes are jointly considered,including local processing,vehicle to infrastructure(V2I)offloading processing and vehicle to vehicle(V2V)offloading processing,and the expression of total task processing delay is given.Then,according to the problem model,under the considering of constraints of computing and spectrum resources and reliability,the optimization problem of minimizing the total delay of task processing is constructed.The original problem is an mixed integer non-linear programming(MINLP)problem,which is NP hard.There is no accurate algorithm for it.In this thesis,the original problem is divided into two subproblems,which is solved step by step.Firstly,the task processing mode of each vehicle is determined,and then the problem of computing and transmission resource allocation in the system is solved.Finally,through the simulation experiment,the influences of the number of vehicles,the total bandwidth of the system and the total computing resources of the edge computing server on the total task processing delay of the system are analyzed,and the effectiveness of the proposed task offloading and resource allocation strategy is verified.