摘要
针对延迟/中断容忍网络特定场景下节点具有的集群运动模式问题,并结合近年来DTN研究领域分簇路由算法的研究进展,提出了基于动态分簇的DTN路由算法。该算法采用基于节点重要度的分簇算法,并选择层次分析法作为节点各参数权重的计算准则;定义节点关联度和稳定度作为普通节点归属特定簇的依据。簇内采取直接递交方式进行消息转发,簇间消息转发时根据节点历史相遇频率选取更可能与目的节点相遇的中继节点。仿真结果表明,与其他经典算法相比,无论是消息生存时间还是仿真时间的影响,该算法在消息递交率和平均延迟等方面都表现出了较好的网络性能。
In view of the issue that nodes in delay/disruption-tolerant network have clustering movement pattern in particular scenarios, and combining the progress of cluster-based routing algorithm in DTN research field for the recent years, this paper proposed a dynamic cluster-based routing algorithm in DTN. It adopted clustering algorithm based on node importance degree, chose analytic hierarchy process(AHP) as the node weight calculation principle, defined node relevance degree and stability degree as the basis of ordinary nodes belonging to a particular cluster. Source node forwarded messages directly to the destination node in intra-cluster, while forwarding messages in inter-cluster, the algorithm selected the node which was more likely to encounter with destination node as relay node according to the historical encounter frequency. The simulation result shows that, compared to other typical DTN routing algorithms, the algorithm can improve the message delivery ratio and reduce the average delay in scenarios with different time to live of messages and simulation time.
出处
《计算机应用研究》
CSCD
北大核心
2015年第11期3395-3398,3412,共5页
Application Research of Computers
基金
陕西省科技计划自然基金重点项目(2012JZ8005)
军事学研究生课题(2011XXXXX-523)
关键词
延迟/中断容忍网络
集群运动模式
层次分析法
历史相遇频率
delay/disruption-tolerant network(DTN)
clustering movement pattern
AHP
historical encounter frequency