Rode Side Unit Collaborative Scheduling And Load Balancing Algorithm In Vehicle Edge Computing System

With the development of intelligent connected vehicles,the vehicle edge computing(VEC)framework has attracted more and more attention.In order to meet the low-latency requirements of on-board applications,the vehicle needs to offload computing tasks to the road side unit(RSU)or the cloud for execution.Due to the uncertainty of the traffic scene,there is a problem that vehicles frequently switch RSU in a short time,or stay in a certain RSU coverage area for a long time,resulting in an increase in the completion time of tasks.In order to minimize this time,this paper analyzes different traffic scenarios from the perspective of adjacent RSU cooperation.The main research work is as follows:(1)Aiming at the problem that vehicles frequently switch RSU when the road is unobstructed,resulting in the increase of task completion time.Using the idea of the distance between the vehicle and the RSU and geometric tangent.an algorithm for solving the remaining travel time of the vehicle within the coverage of RSU communication is designed in the structure of the RSU cooperative communication.The algorithm is used to judge whether the vehicle can receive the calculation result of the task before leaving the coverage of the RSU.If so,the task execution is completed.If not,first sent the calculation result of the task to the adjacent RSU,and then the RSU sends the result to the vehicle.(2)Aiming at the problem that workload of the RSU becomes large when the road is congested,the computing tasks need to be offloaded to the cloud for execution,resulting in an increase in the completion time of the tasks.Using the idea that the communication delay between adjacent RSUs is smaller than the communication delay between RSU and cloud servers,a VEC framework based on RSU cooperation is proposed.In this framework,tasks can not only be offloaded to the current RSU and cloud for execution,but also executed through the RSU collaboration module.In addition,in order to balance the number of tasks in the RSU,a RSU load balancing decision scheme is proposed in the RSU cooperation module.In this scheme,the RSU first judges its own computing load state and task type according to the offloading strategy,and then searches for a suitable RSU to cooperate according to the positioning strategy.(3)Carry out simulation experiments on the above research,the simulation of urban mobility(SUMO)is used to establish a traffic scene where vehicles frequently switch RSU,and the Veins framework is used in OMNet++ to simulate and analyze the remaining travel time algorithm.It can be seen from the simulation results that the algorithm can effectively reduce the completion time of task of the vehicle during the RSU switching process.And use SUMO to build an urban traffic scene including traffic lights and multiple RSUs,and use the Veins framework to simulate the proposed RSU load balancing decision in OMNet++.By comparing the simulation results of the load balancing state of different RSU computing load judgment methods,the effectiveness of the proposed RSU load balancing decision,unloading strategy and positioning strategy is proved.