中文摘要：

由于延迟容忍网络具有高延迟,低数据传输率,拓扑动态变化等特点,设计有效的路由算法一直是延迟容忍网络研究的热点问题.该文在概率模型的DTN中提出了一种数据聚集算法DADTN（Data Aggregation algorithm for Delay Tolerant Networks）.对于目的节点相同的一些数据包可以通过网内聚合（In-network aggregation）方法聚集成为一个数据包.为了衡量每个节点聚集局部数据包和交付数据包到目的节点的能力,算法定义了聚集交付能力的概念.数据包则由聚集交付能力小的节点向聚集能力大的节点进行传输,实施聚集操作.仿真分析表明,与Epidemic和PRoPHET算法相比,DADTN算法能以较低的数据包转发次数获得较高的数据传输成功率.

英文摘要：

Since delay tolerant networks have the characteristics of high latency, low data rate and dynamic topology, designing effec- tive routing algorithms in is a hot issue in the delay tolerant network research area. In this paper, we present a DTN data aggregate al- gorithm--DADTN （ Data Aggregation algorithm for Delay Tolerant Networks） under the probabilistic contact model. The same desti- nation node＇ s packets can be gathered together as a single data packet through the in-network aggregation. In order to measure the data aggregation ability and successful delivery ability of each node, we define a concept of the aggregation and delivery ability. In this al- gorithm , message are delivered form the nodes with low aggregation and delivery ability to the nodes with high aggregation and deliv- ery ability to be aggregated. Simulation analysis showes that compared with the Epidemic and the PROPHET algorithms, DADTN al- gorithm can achieve a higher success rate in the lower number of message forwarding.