论文部分内容阅读
高效的簇路由在分簇过程中不仅要利用车辆的实时信息,更要能够利用车辆间的历史信息来判断其下一刻的链路信息。本文针对道路上车辆速度多变导致的只依据实时信息选取的簇头不稳定问题,提出了一种基于灰色系统理论的分簇算法CA-GST(Clustering Algorithm Based on Gray System Theory)。该算法利用灰色预测模型根据车辆间的历史链路持续时间预测下一刻的链路持续时间,求出车辆间通信链路的连通率。并结合车辆的节点偏差度(车辆的节点度与最佳节点度之差的归一化值)来选举簇头,使得簇头车辆与其成员车辆间链路较稳定,减少了道路上的簇结构个数。最后通过NS2仿真比较了基于灰色预测的分簇路由方法 CA-GST与VMa SC、DMMAC,结果证明本文提出的CA-GST在簇头持续时间、簇头个数、时延、投递率等方面更适应车辆高速移动场景。
Efficient cluster routing in the clustering process not only to make use of real-time vehicle information, but also to be able to use historical information between vehicles to determine the next moment of link information. In this paper, we propose a clustering algorithm based on gray system theory (CA-GST) to solve the problem of cluster head instability, which is only based on real-time information. The algorithm uses the gray prediction model to predict the link duration of the next moment according to the historical link duration between vehicles, and obtains the communication rate of the inter-vehicle communication link. The cluster head is selected based on the degree of node deviation (the normalized value of the difference between the node degree and the optimal node degree) of the vehicle so that the link between the cluster head vehicle and its member vehicles is stable and the cluster structure on the road is reduced Number. Finally, the clustering routing methods CA-GST and VMaSC, DMMAC based on gray prediction are compared by NS2 simulation. The results show that the CA-GST presented in this paper is more effective in terms of cluster head duration, number of cluster heads, delay and delivery rate Adapt to the vehicle speed moving scene.