Research On Task Offloading Optimization Of Internet Of Vehicles Based On Edge Computing

With the continuous development of the Internet of Vehicles,the number of intelligent vehicles continues to increase.In the future,the computing requirements of intelligent vehicle applications will increase greatly.How to meet the computing demand and better reduce the time delay,energy consumption and cost in the process of processing tasks is of great significance to the development of the Internet of vehicles.This research uses mobile edge computing in the Internet of Vehicles as the main technical support,and proposes two multi-task cooperative offloading schemes.The main research tasks of this paper are as follows:Aiming at the problem that the computing power of in-vehicle equipment is limited in the multi-user and multi-server scenario,and the energy consumption and delay need to be optimized,a multi-task and multi-user cooperative offloading scheme based on energy consumption is proposed.Most of the existing solutions only consider the cooperative offloading among equipment,Road Side Unit,and servers,and rarely consider putting cooperative vehicles into the system model.According to the energy consumption and delay required to complete the task,this paper uses the proposed multi-task cooperative offloading algorithm to establish a four-layer task processing architecture of local vehicle,cooperative vehicle,RSU,and Macro Base Station,so as to meet the task offloading needs of more vehicles and better reduce energy consumption and time delay.Through comparative simulation experiments,the performance analysis and comparison of this scheme and other similar schemes show that this scheme can minimize the energy consumption and delay generated by computing tasks.Aiming at the high cost of computing tasks in multi-user and multi-server scenarios,a multi-task and multi-user collaborative offloading scheme based on total computing cost is proposed.Most of the existing solutions put large-scale computing tasks on cloud servers for processing,without considering the cost of such processing and the impact on user Quality of Service.In order to solve this problem,this paper caches the services required for computing tasks on the MBS,uses the Q-learning method to continuously learn the resource allocation scheme of the system,and formulates the optimal calculation offloading strategy.The MBS allocates the cache resources according to the resource allocation scheme.Computation tasks are completed on RSUs with corresponding computing resources.Through comparative simulation experiments,the performance analysis and comparison of this scheme and other similar schemes show that this scheme can significantly reduce the total computing cost of computing tasks.