Vehicle History Trace Based Vehicular Ad-hoc Network Hybrid Routing

Vehicular Ad-hoc Networks (VANETs) is a special kind of Mobile Ad-hocNetworks (MANETs), which achieves vehicle-to-vehicle and vehicle-to-infrastructurecommunication. As an important part of Intelligence Transportation System (ITS),VANETs has wide application scene in many fields, such as vehicle internet access,traffic accident early warning, vehicle automated driving, and attracts extensive focusfrom researchers.Due to the particular features of VANETs, such as the high motilities andnonuniform distribution of vehicles, the network topology changes fast, routing pathsbreak frequently and link capacity is limited, making traditional routing protocolsdeclined seriously. So how to design effective routing protocol is one of the keyproblems in VANETs research works.This paper analyze the characteristics of VANETs and the state-of-the-art routingprotocols in VANETs, and proposes an improved routing algorithm named KPHR(Kalman Predictor-based Hybrid Routing) based on GPSR (Greedy PerimeterStateless Routing). Absorbing the advantages of low end-to-end delay of geographicrouting and high delivery ratio of DTN routing, KPHR can work well in city scenery.In addition, the algorithm uses Kalman predictor to predict real-time location ofvehicles for routing computation, making routing decision more efficient.VanetMobiSim is used to generate the motion trace of vehicles and NS-2is usedto evaluate the performance of research findings. Extensive simulation results showthat the proposed algorithm has better performance in average end-to-end delay,packet delivery ratio and network throughput compared to GPSR and GPSR withbuffer.