车载命名数据网络中的传输机制研究

发布时间：2018-03-01 15:12

本文关键词： 命名数据网络车载网络传输机制中继节点 PIT管理　出处：《重庆邮电大学》2016年硕士论文　论文类型：学位论文

【摘要】：随着现代交通和汽车工业的飞速发展,我国的汽车保有量高速增长,尤其是随着人民消费水平的提高,私家车普及率逐年递增。车载网络作为一种可以提高道路安全性和交通效率,同时满足驾乘人员行车应用需求的车辆互联网络,得到了政府部门与汽车厂商的高度重视。车载网络的特点是节点移动速度快,网络拓扑变化频繁,节点间的连通时间较短,然而传统基于IP网络体系的点对点通信方式难以适应车载网络的动态环境。命名数据网络(Named Data Networking,NDN)作为一种近年来被提出的未来网络架构,可以从根本上改变车载网络的设计方式。作为信息中心网的一个分支,命名数据网络直接应用数据名称来实现数据的分发。不同于当前IP网络架构中将数据发送给某个特定的车辆,命名数据网络则是直接去获取特定的数据,命名数据网络的数据传输模式可以更好的适应车载网络的设计需求。本论文对命名数据网络架构下的车载网络进行了研究分析,并对该架构下的数据传输机制进行了深入研究。首先,针对当前车载命名数据网络普遍以车辆间距作为单一转发节点选择条件,导致多数据块数据回传成功率低的问题,提出一种基于数据属性的Interest包转发机制。该策略综合考虑车辆当前行驶状况,引入多个转发选择条件,并对车载数据属性进行划分,采用三标度层次分析法调整选择条件权重,完成对转发节点的动态选择。仿真分析表明,该策略在保证Interest包传输速度的基础上,提高了Data包回传的成功率,降低了数据获取时延。然后,针对车载网络动态环境中,携带对应PIT(Pending Interest Table)条目的车辆驶离前回传节点信号范围导致Data包回传失败的情况,提出一种动态PIT条目管理机制。车辆处于等待转发阶段时,即使监听到待转发Interest包被其他车辆转发,车辆会根据该机制动态选择是否仍将其添加进PIT列表,并且动态调整PIT条目生存时间,及时清理PIT列表,保证条目数量在合理范围内。此方法在保证PIT条目处于可控范围内的前提下,有效提高Data包回传成功率,增加数据传输的鲁棒性。
[Abstract]:With the rapid development of modern traffic and automobile industry, the number of cars in our country has increased rapidly, especially with the improvement of people's consumption level. The popularization rate of private cars is increasing year by year. As a kind of vehicular network which can improve road safety and traffic efficiency and meet the needs of drivers' driving applications, The characteristics of onboard network are fast moving of nodes, frequent changes in network topology, and short communication time between nodes. However, the traditional point-to-point communication mode based on IP network system is difficult to adapt to the dynamic environment of vehicular network. Named Data networking is a kind of future network architecture proposed in recent years. Can fundamentally change the design of vehicular networks. As a branch of the information center network, The named data network directly applies the data name to realize the data distribution. Different from the current IP network architecture sending data to a specific vehicle, the naming data network is to obtain the specific data directly. The data transmission mode of the named data network can better meet the design requirements of the vehicle network. In this paper, the vehicle network under the architecture of the named data network is studied and analyzed. And the data transmission mechanism under this framework is studied. Firstly, the vehicle spacing is generally used as the single forwarding node selection condition in the current vehicle named data network, which leads to the low success rate of multi-block data return transmission. This paper presents a Interest packet forwarding mechanism based on data attributes. The strategy considers the current vehicle driving condition, introduces several forwarding selection conditions, divides the vehicle data attributes, and adjusts the weight of the selection conditions by using the three-scale AHP method. The dynamic selection of forwarding nodes is completed. Simulation results show that the strategy improves the success rate of Data packet return transmission and reduces the data acquisition delay on the basis of guaranteeing the transmission speed of Data packets. In this paper, a dynamic PIT entry management mechanism is proposed for the failure of Data packet return due to the signal range of the Data return node which carries the corresponding PIT(Pending Interest Table.When the vehicle is waiting for forwarding, Even if it is monitored that the Interest packet is forwarded by other vehicles, the vehicle dynamically selects whether to add it to the PIT list according to the mechanism, and dynamically adjusts the survival time of the PIT entry to clean up the PIT list in time. This method can effectively improve the success rate of Data packet return transmission and increase the robustness of data transmission.
【学位授予单位】：重庆邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN929.5;U495

【相似文献】