| With the rapid development of automobile industry and internet of things technology,VANET as the application of internet of things technology in automobile industry,has attracted extensive attention from academia and industry due to the advantages in traffic optimization,traffic management,traffic accidents avoidance and self driving.The improvement or design of appropriate data transmission protocols is the focus of VANET’s current research,which affects the overall development of VANET.Data broadcast is an important part of data transmission mode in VANET,but rapid change of network topology,unstable communication link,and high requirement of QoS lead to broadcast storm,high packet loss rate,long time delay,network redundancy and other issues.To improve communication performance,many algorithms try to achieve stable communication between vehicles,reduce the channel competition,simplify routing path,optimize the allocation of resources by clustering network,but clusters are not stable and maintain cost is high due to complex characteristics of VANET.In view of the shortcomings of the existing broadcasting algorithms and clustering algorithms,the broadcasting algorithm based on QoS and the clustering algorithm based on intersection are proposed.There are many types of data information in VANET,and different types of data information have different requirements for time delay and reliability.Safety emergency information has high requirements on delay and reliability,audio and video real-time communication information has high requirement only on delay,file information or important status information requires high reliability and has commonly requirements on delay,the delay and reliability requirements of some entertainment application messages are not high.In view of the different QoS requirements of different types of data in VANET,a broadcasting algorithm based on QoS requirements is proposed in this paper.This algorithm takes into account the link quality,the relative distance and the mobility direction to achieve the high reliability and low latency when selecting routing node.Meanwhile,this algorithm is also dynamically adjusted according to the different QoS requirements of data information,so that the data information achieves high QoS for the transmission delay and reliability.Due to rapid change of network topology,unstable communication link,and serious channel competition in VANET,data transmission has high time delay and packet loss rate.Network clustering can improve the stability of inter vehicle communication link,reduce the inter vehicle communication channel resource competition,optimize the allocation of resources and simplify the routing path.But the existing clustering algorithm’s application scenario is single,having network redundancy and other issues.For the unstable cluster in intersection and high maintain cost,clustering algorithm based on intersection is proposed,which taking different clustering methods within and outside the intersection.Outside the intersection,clustering according to the relative speed and relative distance of vehicles,and high similarity mobility vehicles will be divided into a cluster.In intersection,clustering according to turning direction,and the vehicle with the same turning direction will be divided into a cluster,which avoids cluster separation caused by different turning direction.The algorithm takes different clustering methods according to different network scenarios,which is more adaptable to the complex network scenario and improves the stability of cluster well.Finally,we build a VANET system level network simulation platform combined with network simulation and traffic simulation,and then verify the effectiveness of broadcast algorithm based on QoS requirements and clustering algorithm based on intersection.The simulation results show that the broadcasting algorithm based on QoS requirements has better performance than other broadcasting algorithms in terms of time delay and reliability satisfaction.And the clustering algorithm based on intersection is better than the other clustering algorithms in the stability of the cluster,and has more diverse applicable scene. |