Network Technology In 5G Vehicular Ad Hoc Networks

The 5G Ultra Reliable & Low Latency Communication(u RLLC)scenario has greatly promoted the development of vehicle-mounted ad hoc networks(VANET).As one of the most promising areas of the Internet of Things,VANET plays an important role in smart cities,safe driving and in-vehicle entertainment.In VANET traffic scenario,vehicles move at high speeds,resulting in short lifetimes of links between nodes and frequent changes in topology.In order to alleviate this problem,this paper studied on clustering technology and routing protocol.In order to alleviate the problem of topology instability caused by high-speed movement of nodes,this paper proposed a Reliability-and-Stability-based Clustering Algorithm(RSCA).The proposed clustering algorithm first establishes the initial clusters based on the vehicles' location information and driving direction,and determines the node retention by calculating the clustering index of the nodes in the initial cluster.Then it calculates the number of neighbor nodes,average relative speed and link survival time of the nodes in the cluster.The relative reliability of node is calculated based on the link survival time between nodes,and the stability of node is calculated based on the number of neighbors and the average relative speed.Finally,fuzzy theory is used to realize the relative reliability and the stability of vehicle nodes in clusters as fuzzy membership functions to select the appropriate cluster head and the back up cluster head.The MATLAB simulation results show that compared with the random clustering algorithm and the Movement Consistency-based Clustering Algorithm(MCBC),the RSCA algorithm proposed in this paper is more effective in terms of cluster head duration and cluster head change ratio.This paper proposed a Position-based Reinforcement Learning Routing Protocol(Pb RQR)to alleviate the routing instability and data latency problems caused by high-speed movement of nodes and multi-node roting respectively.The proposed routing protocol uses the Q-learnng algorithm of reinforcement learning to learn the topology of VANET.When the vehicle node needs to select the next forwarding node,it needs to consider the rewarding values based on the quality and the location of the node.Vehicle nodes need to maintain and update the Q value table for a long time.It mainly includes two parts,one is that the agent(vehicle node)evaluates the quality of its' neighbor nodes according to the Q-learning algorithm,where the agent iteratively maintains and updates the Q values of its' neighbor nodes in the Q value table.The second is that when the agent needs to forward data,it selects the next forwarding node based on the Q value table and the location information between its' neighbor nodes and the target node.MATLAB simulation results prove that compared with the Ad hoc On-Demand Distance Vector(AODV)and the Greedy Perimeter Stateless Routing(GPSR)protocol,the proposed Pb RQR protocol has better performance in terms of routing hops and duration time.