Research On Task Offloading And Cache Mechanism Of Internet Of Vehicles Based On Edge Computing

The development of 5G technology has led to the rapid construction of intelligent transportation.The emergence of autonomous driving requires huge and fast data exchange,which poses a challenge to the Internet of Vehicles.At the same time,the popularization of electric vehicles also raises new energy consumption issues for the Internet of Vehicles.Mobile edge computing technology is the key to the challenge of the Internet of Vehicles.This paper mainly studies the network architecture of the Internet of Vehicles based on edge computing technology,task processing delay and energy consumption issues,with the purpose of obtaining better quality of experience(Qo E),lower waiting time and lower energy consumption for vehicle users.This paper constructs a framework for the Internet of Vehicles supporting edge computing with a cloud computing center in an urban environment.The framework has a three-layer structure,including the vehicle user layer at the bottom,the edge computing layer in the middle,and the cloud computing center at the top.The main research contents of this paper are as follows:First of all,this article studies tasks with different attributes based on task offloading strategies.Tasks built from entertainment applications do not need strict delay constraints,while tasks built for autonomous driving must have strict delay constraints.This paper divides computing tasks into two categories: delay-sensitive tasks and non-delay-sensitive tasks,and defines basic offloading strategies.This article builds a communication model and a calculation model based on the above two types of computing tasks.The transmission delay expressions and calculation delay expressions of the two types of calculation tasks are obtained respectively.For different task types,based on task processing delay and user quality of experience,the entire network system can obtain different utility values.Finally,the optimization problem of the entire network is modeled as a utility function of the total task processing delay based on the two types of task models.Aiming at this problem,this paper proposes an algorithm for joint optimization of offloading decision and resource allocation decision to maximize system utility.This paper uses an intelligent algorithm ant colony algorithm and a distributed method to solve the problem.The solution process was simulated to verify the effectiveness of the solution method.Then this article studies different types of tasks based on caching strategies.The research on task offloading strategy only studies computational tasks,while the research on non-computational tasks involves caching strategies.When most articles study caching strategies,they only consider the transmission delay of content for non-computational tasks,but this is not in line with the actual situation.This article divides tasks into computational tasks and non-computational tasks.For different types of tasks,the cached content is different,and the task processing process is also different.On this basis,the caching strategy in this paper considers the scheme of moving vehicles,parked vehicles and RSU collaborative caching,as well as the resource allocation strategy of vehicle users and the energy consumption of the vehicle itself.Finally,the problem is modeled as a minimization problem based on energy consumption and task processing delay.This paper uses a distributed algorithm to solve the problem and simulates with MATLAB to show the influence of different conditions and different parameters on network performance.The results of this paper have guiding and reference significance for the establishment of edge computing vehicle network architecture and the selection of network transmission strategies and cache strategies.