Research On Freeway Road Side Unit Deployment Based On Vehicle Clusters

In recent years,the intelligent driving level of automobiles has been continuously improved,but due to the complex driving environment,the safety of autonomous driving systems is still difficult to guarantee.In the Cooperative Vehicle-Infrastructure System(CVIS),the Intelligent Road Side Unit(RSU)has received extensive attention because it can provide road environment information and assist in the networking and communication of the Vehicular Ad-hoc Network(VANET).The main function of the Road Side Unit is to interact with the On Board Unit(OBU)to provide decision information for the vehicle.In addition to RSU functions,RSU deployment has also become a research hotspot.Reasonable deployment of RSU can effectively improve network performance.By adjusting the communication range of the roadside unit,it can adapt to the road environment of different traffic flows and reduce the communication delay and energy consumption of information interaction.This research focuses on freeway RSU deployment,mainly from the following aspects:First,this study analyzes the environmental characteristics and vehicle movement characteristics of one-way four-lane expressways.The composition structure,communication technology and application scenarios of the CVIS are introduced.Introduced the advantages and disadvantages of uniform deployment and hotspot deployment schemes.Secondly,it summarizes several common clustering algorithm classification methods,such as:hierarchical clustering algorithm,density clustering algorithm,and grid clustering algorithm.Analyzed the algorithm principle and advantages and disadvantages of the classical KMeans clustering algorithm in detail.Aiming at the classic KMeans clustering algorithm shortcomings,proposed an improved KMeans clustering algorithm and given the evaluation index of the clustering algorithm,such as network residual energy and number of surviving nodes.Based on the above research,an RSU uniform deployment scheme based on vehicle clusters is proposed,that is,the RSU deployment distance is the average vehicle cluster length plus twice RSU communication radius.The network connectivity rate and end-to-end delay are taken as the performance indicators of RSU deployment effect.Then,detailed communication environment,road environment and vehicle simulation parameters are set for the research object.The NS-2 simulation software was used to analyze the impact of RSU,vehicle communication radius and vehicle speed on vehicle-road communication.Using MATLAB software,compare the improved KMeans clustering algorithm with the classic KMeans clustering algorithm in terms of the relationship between the number of cluster heads and energy loss,the distribution of vehicle clusters,the number of surviving nodes,the network remaining energy,and end-to-end delay.The simulation results show that the improved KMeans clustering algorithm has significantly better clustering effect than the classic KMeans clustering algorithm.Finally,based on the improved KMeans clustering algorithm,analyzed the relationship between the RSU communication radius and vehicle density,network connectivity,and average vehicle cluster length when used the RSU deployment scheme proposed in this study.This provides a basis for setting the RSU deployment interval and communication radius.