Research Of Broadcast Technology In Vehicular Network

In recent years, with rapid development of economy, science and technol-ogy and automobile industry, road safety and traffic efficiency are two crucial problems in transportation systems. Consequently, Intelligent Transport Sys-tem (ITS) arises. Vehicular Network, as an important component of ITS, plays a very important role in improving road safety and traffic efficiency.The characteristics of Vehicular Network can be summarized as high-mobility nodes, highly dynamic network topology, poor link quality, long propagation delay. These characteristics make broadcast protocols become the most ef-fective method in the distribution of traffic information of Vehicular Network.However, the change of vehicle velocity, density and network load in road traf-fic will largely affects broadcast performance of Vehicular Network. Exist-ing research of broadcast technology in Vehicular Network still can not adjust broadcast scheme adaptively based on change of network conditions, resulting in frequent occurance of broadcast storm, unreasonable allocation of network resources and insatiable Quality of Service (QoS) requirements of different in-formation.To cater to the requirements of broadcasting in Vehicular Network and improve broadcast efficiency, this paper investigates broadcast technology of Vehicular Network based on different traffic condition under two scenarios V2V and V2I, especially analysing broadcast performance from the following three aspects.Firstly, QoS and queue management are critical issues for broadcast scheme in Vehicular Network. Lack of model and analysis of queue in MAC can not obtain important QoS performance accurately and then can not know protocol performance well. This paper presents a 2-dimensional (2-D) Markov chain for analyzing QoS performance of realistic 802.11 broadcast systems with finite buffer based on different network load and network size under V2V scenario.Also, this paper derives an simplified method for solving the steady state prob-abilities of the Markov chain. Our analyses reveal that the lack of binary ex-ponential backoff and retransmission in the 802.11 system results in poor QoS performance during heavy traffic load, particularly for large Vehicular Network.Such performance deterioration can be avoided by proper traffic control which provides traffic control guidelines to maintain good QoS performance for Ve-hicular Network.Next, in Drive-thru Internet, because of different locations to road-side units (RSUs), vehicle nodes have different transmission rates. If vehicle nodes can not change frequence of sending packets adaptively based on change of transmission rates, the system performance is throttled to the minimum trans-mission rate, resulting in low system throughput. With regard to this problem in Drive-Thru Internet, as there is not still an efficient MAC scheme for solving the problem, this paper presents a novel MAC scheme for Drive-Thru Internet in a sparse highway environment and designs a Markov chain model to anal-yse the performance of new broadcast scheme. Based on the model, it is shown that when vehicle number is small the proposed MAC scheme can obtain higher throughput and mitigate the impacts of vehicle mobility on the system through-put. Extensive simulations validate the accuracy of the analytical model and effectiveness of the proposed MAC scheme.Finally, in practical radio transmissions, capture effect is a dominating fac-tor that affects wireless network performance. Lack of model and analysis of capture effect can not abtain evaluation on network performance, especially in the highly mobile vehicular environment. This paper presents a performance-prediction model for analyzing performance of Drive-Thru Internet based on capture effect under different moving speed and network size. Using a vehic-ular traffic flow model models vehicular movement on road and describes the relationship among velocity, density and traffic flow, which accurately simu-lates traveling situation of vehicles. Using the performance-prediction model,optimal contention window value can be obtained, by which the best system throughput can be reached without wasting contention time.