基于NDN架构的车联网内容检索协议

在已有的基于NDN架构的车联网（vehicular named data networking，VNDN）内容检索协议中，对于地理位置无关应用的内容检索通常是以洪泛方式获得内容提供者的位置，这种方式不仅开销大且不能适应内容提供者位置的高度动态变化.本文提出了一种基于分簇和预测机制的VNDN内容检索协议CPCoR （content retrieval protocol based on clustering and prediction mechanism）.CPCoR通过车辆和路边单元的协同交互，建立了动态的网络内容索引表，在街道内转发机制中选择链路稳定性较好的簇头进行转发，在街道间转发机制中利用街道的路由性能来选择最优的转发街道.实验结果表明，与已有的Navigo和CCVN协议相比，CPCoR协议提高了21.5%和51.3%的内容获取成功率，且有效降低了22.6%和28.4%的内容获取成本. 

Content Retrieval Protocol of Vehicular Network Based on NDN Architecture

In the existing content retrieval protocols of vehicular named data networking (VNDN), the content retrieval for location-independent applications is usually to obtain the location of the content provider based on flooding. This method not only costs much, but also is unable to adapt to the highly dynamic changes of the location of content providers. In this paper, a content retrieval protocol, CPCoR, was proposed based on clustering and prediction mechanism for VNDN. The CPCoR was arranged to establish a dynamic network content index table through the collaborative interaction between vehicles and roadside units. In the intra-street forwarding mechanism, the index was designed to forward according to the cluster head with stable link connection. In the inter-street forwarding mechanism, the index was designed to select the best forwarding street according to the routing performance of streets. Experimental results show that, compared with existing protocols, Navigo and CCVN, CPCoR protocol can not only improve the success rate of content acquisition by 21.5% and 51.3%, but also can effectively reduce the cost of content acquisition by 22.6% and 28.4%.