The Research Of Checkpoint Server Selection Method Based On Autonomous Vehicular Cloud

With the development of wireless communication technologies and the performance of the car equipment, the conception of Autonomous Vehicular Cloud (AVC) is proposed. AVC can provides many resources and services,but it has short communication time in the network, loses the allocated resources easily, and leads to waste recourses due to the high mobility of vehicles on the highway. This paper endeavors to solve the problem of inefficient aggregation and allocation of resource of vehicle nodes under the highway scene. Based on the specific of AVC on the highway, this paper introduces the checkpoint server mechanism, and proposes three checkpoint server selection methods. They are selecting the node near the task initiator, selecting the node near the task bearer and selecting the node randomly. In the process of aggregation and allocation of resource, it allows Checkpoint server to save resources that have been allocated or collected. When needing the resources, it can quickly and efficiently provide. So it can improve the efficiency of aggregation and distribution of resource.This paper uses OPNET network simulation tool to achieve three options, analyzes and verifies the results in different vehicle node density scene. Eventually this paper concludes that selecting the node near the task initiator as a checkpoint server which can improve efficiency of aggregation and obtain obvious performance in the case of the sparse vehicle nodes.