Research On Multi-Task Slice And Allocation Strategy Of Autonomous Vehicular Cloud In The Highway

With the improvement of the performance of mobile intelligent terminal and the development of wireless communication technology,the function and performance of on board unit have been improved,and the vehicle can act as a service provider.Therefore,a request composed of a plurality of tasks in the vehicle can be assisted by other vehicle nodes to perform the task in the Autonomous Vehicular Cloud(AVC).When a user’s services that consist of multiple tasks,tasks have dependencies and are connected into a service sequence according to the order of the task dependencies.Taking into account the stability of the node communication link and its computing capability,the tasks connect in the order of dependencies into a service sequence accordingly,minimizing the task completion time to deploying the task,thereby to ensure the smooth execution of services and improve the success ration of the task execution.Therefore,this thesis studies the problem of task deployment and allocation based on service sequence to improve the success ratio of tasks and the utilization efficiency of node resources.The work of this thesis is divided into task deployment phase and task allocation phase.In the task deployment phase,the available resources to execute task within the deadline are predicted according to the vehicle movement information,and the task is deployed to the roadside unit by minimizing the total task completion time based on the resource prediction task completion time.In the task assignment phase,roadside unit dynamically divides the task into multiple subtasks and assigns them to the vehicle nodes according to the workload of the task and the vehicle resources,so as to ensure the utilization of the vehicle resources and the success ratio of the tasks.In the thesis,optimized performance network engineering tool(OPNET)is used to verify the performance of the task deployment and allocation scheme based on the services sequence in different task types and different workloads.The scheme is also compared with the optimization generation scheduling(OGS)algorithm,classic task scheduling algorithm,which shows the success ratio of this scheme is 11.36% higher of OGS algorithm for the dependency task in the highway AVC with rapidly changing topology.