An Edge Computing-based Autonomous Vehicle Group Cooperation Model in a Highway Scene

Most existing research on autonomous vehicle groups mainly focuses on semi-centralized network control, without considering computational difficulties caused by the increasing amount of data. Under external inherent disturbances, it is difficult for autonomous vehicle groups to realize cooperative computing. To cope with these problems, an autonomous vehicle group cooperative model in a highway scene based on edge computing is given. First, a new vehicle-to-vehicle computation offloading method is defined and a hierarchical multi-role autonomous vehicle group formation method is given based on it. Next, the evolution events of vehicle groups are investigated and a dynamic evolution algorithm is proposed. Finally, an autonomous vehicle group cooperation model is proposed by considering the reliability, cooperativeness, punctuality, stability, and evolvability of an autonomous vehicle group. Simulation results show that the proposed method can form higher-performance vehicle groups than the existing methods.