基于历史轨迹预测的车辆自组织网络混合路由算法

发布时间：2018-01-26 23:49

本文关键词： 车辆自组织网络卡尔曼预测地理位置路由 DTN路由　出处：《天津大学》2014年硕士论文　论文类型：学位论文

【摘要】：车辆自组织网络(Vehicular Ad-Hoc Networks, VANETs)实现了车辆-车辆间通信与车辆-路边设施间通信，是智能交通系统(Intelligent Transport System，ITS)的重要组成部分，在车辆移动Internet接入、交通信息预警、车辆行驶辅助等方面具有广阔的应用前景，成为近年来的研究热点。作为一种特殊的无线移动自组织网络(Mobile Ad-Hoc Networks, MANETs)，车辆自组织网络其独有的特点，如车辆节点运动速度较快、车辆节点分布不均等，，造成节点之间通信时间短、链路频繁断裂和链路容量受限。这些问题给设计车辆自组织网络路由协议，实现高效数据传输带来了极大的挑战。本文在分析车辆自组织网络特点的基础上，对现有的车辆自组织网络路由协议进行了深入的研究和分析。针对现有协议的缺陷，提出了一种基于卡尔曼预测的混合路由算法(Kalman Predictor-based Hybrid Routing, KPHR)。该算法以GPSR(Greedy Perimeter Stateless Routing)协议为基础，结合了地理位置路由在连通性较好网络中低延迟、时延容忍网络(Delay Tolerant Networks, DTN)路由在连通性较差网络中高投递率的优点。同时分析现有车辆自组织网络路由协议中所采用的获取车辆位置方法的不足，利用卡尔曼预测器对车辆实时位置进行高精度的预测，使用预测位置参与路由计算，辅助地提高路由决策的效率。本文通过分组平均端到端时延、分组投递率和网络吞吐率三个方面将KPHR同GPSR和带有缓存的GPSR进行了比较和评价。利用真实地图数据，通过VanetMobiSim构建道路拓扑用于协议仿真，使仿真场景更加贴近实际。将得到的仿真场景用于NS-2仿真平台，获得仿真数据。仿真结果表明，在分组平均端到端时延、分组投递率和网络吞吐率方面，KPHR路由算法具有更好的性能。
[Abstract]:Vehicle Ad Hoc Ad-Hoc Networks (VANETs) implements vehicle-vehicle communication and vehicle-roadside facility communication. It is an important component of Intelligent Transport system (ITS) in the vehicle mobile Internet access. Traffic information early warning and vehicle driving aid have a wide application prospect and become the research hotspot in recent years. As a special wireless mobile ad hoc network, Mobile Ad-Hoc Networks, Manets has its unique characteristics. Such as fast moving speed of vehicle nodes, uneven distribution of vehicle nodes, short communication time between nodes, frequent link breakage and limited link capacity, these problems lead to the design of vehicle ad hoc network routing protocol. The realization of efficient data transmission brings great challenges. Based on the analysis of the characteristics of the vehicle Ad hoc network, this paper makes a deep research and analysis on the existing routing protocols of the vehicle ad hoc network, aiming at the defects of the existing protocols. A hybrid routing algorithm named Kalman Predictor-based Hybrid Routing based on Kalman prediction is proposed. The algorithm is based on the GPSR(Greedy Perimeter Stateless routing protocol. It combines the geographical location routing with delay Tolerant Networks in the well-connected network with low delay and delay tolerance. DTN-based routing has the advantages of high delivery rate in poor connectivity networks. At the same time, it analyzes the shortcomings of the existing routing protocols used in the Ad hoc network to obtain the location of vehicles. The Kalman predictor is used to predict the real-time position of the vehicle with high precision, and the predictive position is used to participate in the routing calculation, which can improve the efficiency of routing decision. In this paper, KPHR is compared with GPSR and GPSR with cache in terms of average end-to-end delay, packet delivery rate and network throughput. Real map data are used. The road topology is constructed by VanetMobiSim for protocol simulation, which makes the simulation scene more close to reality. The simulation scene is used in NS-2 simulation platform. The simulation results show that the KPHR routing algorithm has better performance in terms of packet average end-to-end delay packet delivery rate and network throughput.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U495;U463.67

【参考文献】