Research And Implementation Of Clustering Optimization Algorithm For VANET Based On Mobility

Nowadays,with the rapid social and economic progress and the rapid popularization of automobiles,urban traffic safety problems and traffic congestion problems are becoming more and more serious.In order to provide a more comfortable and safe transportation experience,domestic and foreign research organizations attach great importance to the research of Internet of Vehicles and related technologies.As a typical application of mobile self-organizing network in vehicle transportation,car networking is the core of smart city intelligent transportation system.It uses on-board equipment to obtain information for rapid communication between people and vehicles,vehicles and vehicles and vehicles and equipment,so as to realize the city Intelligent integrated network.However,due to its characteristics such as strong node mobility,rapid network topology change,and node movement trajectory restricted by roads,which are different from mobile self-organizing networks,the typical and effective mobile self-organizing network clustering algorithm in the past is not suitable for Internet of Vehicles.In this case,it is urgent to design a clustering algorithm that is more in line with the characteristics of the connected car city scene,and that is efficient,stable,and reliable.This paper deeply studies and analyzes the network characteristics of the Internet of Vehicles and the existing clustering mechanism,and conducts the following research on the performance of cluster stability,network overhead,and latency:First,a mobility-based VANET clustering algorithm is designed.Through three parameters(vehicle movement direction,average vehicle speed,and inter-vehicle distance)to select a stable cluster head,an optimal clustering scheme can be found.The algorithm is simple and stable enough to perform relatively few handovers and cluster reorganization.The number of packets and the delay show that a stable cluster has been achieved in a relatively long time.Secondly,in view of the poor stability of the cluster head in the cluster maintenance phase of the previous algorithm,a VANET clustering optimization algorithm based on mobility weights is proposed,which increases the number of neighbor node connection parameters in the mobility information.An auxiliary cluster head is introduced to assist the main cluster head to manage the cluster during the cluster maintenance phase to ensure a smooth transition to the maintenance phase to meet the strict requirements of VANET environment security applications.Finally,in order to optimize the previous algorithm,when re-clustering occurs at the intersection,it will not only increase the clusterpacket cost,but also greatly increase the number of clusters in the system.An adaptive intersection based on extreme learning machine is proposed.Clustering algorithm,which has the ability to adapt to the behavior near intersections and prepare clustering for intersections to mitigate the effect of re-clustering,which can achieve more stable and effective clustering.