Research On The Clustering Routing Algorithm For Reliable Communication In VANETs

As an important part of intelligent transportation system,vehicular ad-hoc networks(VANETs)provide flexible and fast data transmission support for solving key problems such as road traffic efficiency and vehicle safety driving.However,due to the high mobility of vehicles and the randomness of driving trajectory,the topology of VANETs presents transient and volatile.At the same time,the difference of road driving conditions also brings about the uneven distribution of network nodes,which brings severe challenges to the reliable transmission of data in VANETs.For an Ad Hoc network with a certain scale such as VANETs,the clustering methods can effectively improve the stability of the network,thereby contributing to the reliable transmission of data.In view of this,the dissertation proposes a new global affinity propagation clustering algorithm.Based on the obtained clustering structure,a reliable clustering routing algorithm based on reinforcement learning is proposed.The main research work of this dissertation is as follows:1.Firstly,by introducing communication-related parameters,the similarity function of the affinity propagation algorithm is reconstructed,so that the vehicles with low relative mobility and high communication performance are more likely to become cluster heads.Secondly,three scaling functions are formally defined to quantitatively assess the effect on the cluster stability when a vehicle joins it.Then,from the perspective of global competition,a cluster head selection mechanism under the condition of multiple cluster heads is designed.Finally,for the vehicles in three states,combined with the process of cluster head selection,cluster formation and cluster maintenance,the process of vehicle state transition is given,and a new global affinity propagation clustering algorithm is proposed.The Simulation experiments for different traffic scenarios show that the algorithm is generally better than existing similar algorithms in the cluster stability.2.Based on the stable cluster structure obtained by the proposed clustering algorithm,firstly,the cluster head vehicles and the gateway vehicles are selected as the relay vehicles for data cross cluster transmission to reduce the forwarding amount of route request messages during the route establishment process,so as to avoid the occurrence of broadcast storm problem.Secondly,by introducing evaluation factors such as link duration,link communication rate,direction factor and congestion factor,the reward function of reinforcement learning algorithm is reconstructed to improve the survival time and communication quality of the route.Finally,from the perspective of Q table structure update and Q table value update,the realization processes of route establishment and route maintenance are given respectively.At the same time,in order to improve the reliability of data transmission and ensure the effectiveness of the established route,according to the idea of "discovering while transmitting",a dynamic route establishment process is given,and a reliable clustering route algorithm based on reinforcement learning is proposed.The simulation results for different traffic scenarios show that the algorithm has better communication performance than other similar algorithms.In summary,the research work and achievements on reliable communication of VANETs not only meet the actual data transmission requirements of vehicle to vehicle(V2V)communication in VANETs,but also further enrich the methods and theories of clustering and routing of VANETs,which has important theoretical and practical significance for supporting intelligent transportation system to effectively solve various traffic problems.