| China is the largest vehicle production and consumption country in the world, while vehicle has become one of the most important parts of cities. With the de-veloping of economy, city resident population is increasing constantly. Traffic con- gestion, vehicle accidents, environmental pollution has become a daily condition which is an urgent problem to be solved in metropolis. At present, all the countries are seeking the solutions to improve transport efficiency and promote transport system to the intelligent evolution. For this purpose, Internet of vehicles(Io V) technology emerges. Io V can make vehicles realize the functions including real time traffic information acquisition, route planning, lane-change warning, collision warning, internet access and various other intelligent driving services, which can improve the traditional driving concept and further enhance the traffic efficiency base on the existing transport infrastructure. In short, Io V merges wireless com-munication technology with vehicle engineering and traffic engineering. It can forward the development of traditional industry by environment sensing, infor-mation interaction and information integration.Io V is a type of wireless communication network with predefined protocol and data exchange standard. The academic field mainly divides this technology into two categories, which are communication between vehicle and roadside infra-structure(V2I) and communication between vehicles(V2V). For the features that V2 I communication allows vehicles acquire traffic information from roadside unit(RSU), it can provide corresponding driving guidance for each terminal from the macroscopical vision.Therefore, V2 I communication has more advantages in ser-vice categories than V2 V communication. For the aforementioned issues, V2 I communication is a key research area, and it attracts much attention in intelligent transportation field. Base on the above considerations, the research in this disser-tation mainly focus on V2 I communication.Media access control(MAC) sublayer in wireless communication supply the function of channel access control, and it also achieve the function of control for physical(PHY) layer. Therefore, the performance of MAC protocol impacts on not only channel resources scheduling, but also quality of service(Qo S) guarantee. IEEE 802.11 p protocol is adopted is the PHY and MAC layers of V2 I communica-tion. However, the original intension of IEEE 802.11 series of protocols is to providide best effort services for indoor users with low mobility. If apply this protocol into vehicular communication field, one of the most important problems in MAC layer should be considered is that it cannot provide Qo S guarantee. Aim-ming at the above issues, several key technologies in the MAC layer of V2 I com-munication is investigated in this dissertation. The research mainly focuses on the rate adaption, MAC performance modeling for multi-user and multi-rate scenario and extended problem of RSUs seamless deployment, Qo S guaranteed adaptive admission control for safety-related services under non-seamless covered scenario.As the theoretical principle of this dissertation, the key technologies about PHY and MAC layer of IEEE 802.11 p is introduced at the beginning. Aimming at PHY layer, the air interface parameters of orthogonal frequency division multi-plexing(OFDM) based multi-carrier modulation and coding scheme(MCS) is presented. Besides that, the relationship between channel condition and network performance including bit error rate(BER), packer error rate(PER), transmitting rate is investigated. For MAC layer, several important frame structures and chan-nel access schemes are introduced. Due to the reason that the research work in this article is mainly founded by simulation, the accuracy of radio propagation model is much critical. However, traditional researches of propagation models for IEEE 802.11 mainly focus on low mobility terminals. When applying the protocol in V2 I communication, frequency selective fading, multipath transmission and Doppler shift will result in additional affectes on signal transmition. Therefore, field test for radio transmitting features are presented for V2 I communication. Experiment results indicate that log-normal transmition model can reflects the radio propaga-tion feature. This experiment supplys theoretical supports for the simulations in the rest of this dissertation.IEEE 802.11 p supports multi-rate in PHY layer to adapt different channel condition. However, rate selection and switching scheme is not specified in the MAC layer of the protocol. Generally, signal is fast time-varying is V2 I commu-nication scenario. Besides that, dwelling time of vehicle is relative short limited to its high mobility and the small coverage of RSU. Received signal strength(RSS) will appear climbing trend in a short period and then turn into declining trend during the process that vehicle passes by RSU. Aimming at this problem, a rate control algorithm with fast channel response feature is proposed. It solves the tra-ditional problem of the balancing between channel response speed and transmition reliability in V2 I communication. Simulation results in the aspects of throughput, BER and PER validate the proposed algorithm performs better than existing tradi-tional algorithms in V2 I communication.Network performace directively influenced by the amout of associated ter-minals according to the share-channel scheme in the MAC layer of IEEE 802.11 p. Considering the financial issue, network planning for V2 I communication should utilize the limted spectrum resources effectively in the premise of guaranteeing Qo S. Therefore, RSUs deployment issue has important practical significance in V2 I communication. It should allow vehicles access seamlessly, meanwhile Qo S of safety-related services should be guaranteed. Aiming at this problem, mathe-matical model of network performance under multi-user and multi-rate scenario is derived for V2 I communication at first. Base on the aforementioned model, the theoretical boundary of safety-related services Qo S guaranteed access radius is investigated under maximum traffic condition. Further, a safety-related services oriented seamless RSU deployment scheme, which solves the problem of Qo S guarantee in V2 I communication, is proposed in this dissertation.Due to the financial issue, RSU cannot be deployed in the non-hot-spot area of cities, and V2 I network cannot compose seamless coverage. For an individual RSU, network load will ascend sharply with the increasing of traffic density. Therefore, admission control should be considered in non-seamless covered V2 I communication network. Aimming at this issue, a fuzzy Q-learning based admis-sion control algorithm for V2 I communication is proposed. It solves the problem that Qo S of safety-related services cannot be guaranteed in non-seamless covered scenario. The proposed algorithm has adaptive learning capability, and prior knowledge such as propagation model, network parameter, and traffic parameter is not a necessity for the proposed admission control algorithm, which has outstand-ing universality. Simulation results validate that the proposed algorithm can bal-ance the coverage and network capacity, and access non-safety-related services as much as possible in the premise of guaranteeing the Qo S of satety-related services, which can sufficiently utilize the spectrum resources. |
PDF Full Text Request