Research On The Performance Of The Services Delay In The Internet Of Vehicles System Based On Fog Computing

As an important embodiment of smart city in intelligent transportation,the Internet of vehicles is an important application of the Internet of things,which is recognized as one of the most important development directions of the future communication networks.Owing to the imperfect roadside computing system,the Internet of vehicles is currently mainly used in information services,while less attention,most of which is concentrated on accident first aid,is paid to road safety.Thanks to the rapid development of the Internet of things,with more and more devices connected to the network,the fog computing capability of the Internet of vehicles is increasing,making the internet of vehicles possible to break through to a new stage of development.In order to realize the ultimate goal of Internet of vehicles,it has been the primary task to establish a complete computing and communication system of the Internet of vehicles.It presents a new system model based on fog computing to deal with autonomous driving service and studies the scheduling problem from the perspective of service delay in this paper.In this paper,combined with the main wireless communication environment of the vehicles,it proposes a new system model for processing autonomous driving services in internet of vehicles under the fog computing framework,the main characteristics of which are introduced from three aspects: network model,service model and delay model.From the perspective of services processing,the system is mainly divided into three processing layers,including user layer,edge layer and core layer.The user layer is mainly composed of local vehicles,to improve the efficiency of which the size of vehicles ad hoc networks and the allocation of the services need to be studied.Clustering can improve the efficiency of the edge layer which is mainly composed of roadside intelligent devices,and it need to be determined the number of clustering and clustering algorithm.The core layer is mainly composed of cloud data centers,and its efficiency depends on the routing distance between users and the core layer.Autonomous driving services are allocated to three layers for balanced processing to get the fastest response,and the question of the services allocation needs to be solved.Based on the proposed system model,in order to determine the business allocation ratio and minimize the delay of a single autonomous driving service,it verifies the results from two perspectives of mathematical derivation and performance simulation in this paper.It shows that the user layer has the highest efficiency with the optimal combination of the size of vehicles ad hoc networks and services allocation ratio,and the optimal scheme is affected by the number of users.The optimal number of clusters in the edge layer is the trade-off between communication delay and computational delay.The optimal services allocation ratio in the three processing layers is affected by the efficiency of each processing layer.In addition,it introduces a genetic algorithm to cluster the edge layer in this paper,and the results show that the algorithm can rapidly converge to the optimal solution of the trade-off between communication delay and computation delay under security constraints.What's more,the algorithm has good adaptability.