Research On VANET Routing Algorithm Based On Fuzzy Logic And Reinforcement Learning

In recent years,with the continuous improvement of people’s living standards,car has become the main means of transportation of people.Vehicular Ad Hoc Networks(VANET),combining cars with wireless communication,is about to become an important communication method.The core vehicle can collect and process information carried by all vehicles in the interesting area by multi-hop routing.However,due to the special characteristics of the vehicle with high moving speed and restricted street direction,the traditional wireless ad hoc network routing protocol will no longer be suitable for the VANET.Therefore,this thesis studies a routing algorithm suitable for VANET.First,this thesis introduces two simulation tools of VANET,Simulation of Urban MObility(SUMO)and Network Simulation-2(NS-2).SUMO is a urban streets simulation software that simulates randomly moving vehicles on urban roads and introduces actual maps and traffic rules.SUMO can export simulation results for simulation tools,such as NS-2 or MATLAB.We studied three traditional MANET routing protocols AODV,DSDV,DSR.We used the NS-2 to compare the packet loss rate,throughput,and delay performance of these three routing protocols at different node moving speeds.And through analysis,it is concluded that these three routing protocols are not applicable to VANET.Secondly,three routing algorithms based on group intelligence optimization are introduced,such as ant colony optimization algorithms,particle swarm optimization algorithm and genetic optimization algorithm.These three routing algorithms are applied to a random mobile network scenario.All the three routing algorithms are iterative routing algorithm.Through multiple iterations,the next hop node that is most suitable for forwarding data packets is found.The performance of the group intelligent optimization algorithm is verified by MATLAB simulation and compared with the traditional AODV routing algorithm.The group intelligent optimization algorithm can effectively reduce the number of routing hops.Among them,the ant colony optimization algorithm has the lowest complexity.However,this ro uting algorithm cannot be updated in time according to the network topology changing.It is not applicable to VANET multi-hop transmission.Finally,this thesis introduces the fuzzy logic algorithm.The algorithm can comprehensively consider multiple factors such as mobility,follower density,bandwidth efficiency,etc.to evaluate the suitability of neighbor nodes as the next hop.The fuzzy logic algorithm is introduced into VANET to establish an optimal route for the communication between two points.In addition,in order to minimize collisions,a cluster forwarding routing algorithm based on fuzzy logic is used to improve the connectivity of the entire network.Considering the complexity of the algorithm and the average number of hops,this thesis uses reinforcement learning to optimize the first and last two hops to reduce the average delay of the network.